From dkelly Wed Aug 28 13:45:07 2002 To: dlp@ipac.caltech.edu Cc: alberto@ipac.caltech.edu, chad@as.arizona.edu, dhines@as.arizona.edu, dkelly@as.arizona.edu, jkeene@sirtfweb.jpl.nasa.gov, latter@ipac.caltech.edu Subject: weeks 7-8 campaign structures Hi, The following discussion covers the proposed sequence of MIPS tasks for weeks 7-8, campaigns H,I,J,K. I also include a discussion of the standard calibration sequence to be used in SV and throughout the mission. Dean and Jocelyn have the most schedule-sensitive tasks, so they ought to review my logic in putting together these sequences. Anyone else who feels like reading 400 lines of text can do the same. Alberto usually takes responsibility for implementing these sequences, but I think the job now falls on Debbie. Please forward this to the appropriate person if that is not the case. Doug --------------------------------------------------------------------- MIPS CAMPAIGN H time since Start-up activities, Si and Ge MIPS-905 35 Ge anneal Routine 24um flux standard MIPS-920 11 24 Scattered background monitor MIPS-950 30 35 Routine 24um flux standard MIPS-920 11 65 160um First light and Ph/SR AOT val.(Prel)MIPS-324 18 76 160um Stim response MIPS-248 10 94 Routine 24um flux standard MIPS-920 11 104 IOC Darks, 160um MIPS-912 9 115 IOC Darks, 24um MIPS-910 15 124 Routine 24um flux standard MIPS-920 11 139 IOC Darks, 70um MIPS-911 30 150 160um Voffset adjustment test MIPS-127 5 180 Routine 24um flux standard MIPS-920 11 185 70um Spectral leak check MIPS-241 50 196 Ge Anneal and recovery MIPS-930 12 246 160um On-board data processing algorithmsMIPS-112 15 10 Calibration stars against different bkgdsMIPS-921 60 25 160um IC - routine, Phot AOR MIPS-919 21 85 70um IC - routine, Phot AOR MIPS-918 21 106 Periodic IRU Calibration updates PCS-055 36 127 Periodic PCRS/STA alignment filter updatesPCS-105 12 163 70um WF Focal plane survey (coarse) MIPS-132 126 175 24um Flat - routine, Phot AOR MIPS-917 10 301 24um Flat - routine, Phot AOR MIPS-917 10 311 24um Scan AOT validation (Preliminary) MIPS-326 18 321 24um Frame table update sanity check MIPS-960 8 339 Shutdown activities, Ge and Si MIPS-992 19 347 Comments: The driving factor at the start of this campaign is the multiple runs of MIPS-920. These 5 repeats are designed to study the stability of the 24um array response over the first three hours after instrument power-up. To accomplish this, we need to run MIPS-920 right after MIPS-905 and then again at semi-regular intervals over the next several hours. The other big influencing factor is the need to schedule 227m of 70um, 78m of 160um activities, and 90m of tasks covering all 3 arrays (MIPS-950 and MIPS-921) within the windows of good Ge operations following the anneals. The Ge anneals occur fairly early during the startup activities, further restricting the available Ge test time. The result of all this is a 3-part campaign structure, with a mix of 24um responsivity checks and Ge tasks during the first part, with the rest of the Ge tasks and the FPS in the second part, and with the remainder of the 24um tasks in the third part. In ground testing, the 24um responsivity changes were greatest during the first hour, so I tried to schedule short activities in between the early runs of MIPS-920 to get good sampling. I also wanted to minimize pointing overheads by linking together tasks that require identical pointing and by emphasizing observations near the CVZ. I put the scattered background monitor test first, in great part to provide consistency with earlier campaigns. Also, this is the last execution of this task and the only one at nominal temperatures, so it is important that we get good Ge measurements. I put the 160um first light and stim response tasks next because I think it is essential that we get those tasks done early, before the 160um responsivity has changed much. I am mainly guarding against stim saturation. If we want to be overly cautious, we could schedule this task first after MIPS-930, when only 10m of cosmic ray damage has occurred. I think the better approach for the start of IOC is to select a stim brightness that avoids saturation problems. The 160um and 24um darks were lumped together to avoid pointing overheads, with the 160um measurement coming first for cosmic ray reasons. I put the 160um Voffset test right after the 70um darks since the pointing is the same. I put the 70um spectral leak check last during this first part of the campaign. There will be 4h of radiation damage on the array by the end of this task, which is large but not excessive. We plan to look at bright sources and to look in detail at the PSF. Saturation during this test would be a disaster, so care is needed in the source selection. In the second part of the campaign, I open with the on-board data processing task. It is performed while observing a standard star, which is why I paired it with MIPS-921. These tasks are placed ahead of the illumination correction measurements to put the IC measurements the same amount of time after the anneal as they will be during standard startup activities. This places MIPS-921 early after the anneal, when the stim/source calibration at 70um is potentially not so good. The guard against this is to put the 70um observations from MIPS-921 towards the end of that task. The IC measurements are made together since they are measured far from the CVZ. This puts the start of the 70um FPS at 2h after the anneal and the end of this task at 4h after the anneal. These times are again large but not excessive -- we should expect the responsivity to be up about 1.5-2x during this test. Saturation would mess up the centroiding, so we need to be careful in the source selection. With the FPS starting 6h after the start of the campaign, we can expect good temperature stability for the telescope and electronics, which improves the pointing stability. As for the PCS tasks, the IRU calibration has to happen every 24h and the PCRS calibration has to take place every 8h. There is only +/- 1h of tolerance on these times. Working from the current timeline, that puts these tasks where I list them above, right before the FPS. That would delay the start of the FPS by 48m, which is starting to get a bit uncomfortable. It would be a lot nicer if we could defer the PCS commands until after the FPS. There is no particular logic behind the ordering of the tasks in the third part of the campaign other than putting the two 24um flat observations together to reduce pointing overheads. This campaign involves the first 160um operations and the last scattered light measurement. It will be difficult to do much at 160um if the telescope temperature is above 7K. The scattered light test should possibly be deferred if the telescope is not yet close to its final temperature. --------------------------------------------------------------------- MIPS CAMPAIGN I Periodic PCRS/STA alignment filter updatesPCS-105 12 time since Start-up activities, Si and Ge MIPS-905 35 70um anneal 70um and 160um Optimum bias setting MIPS-246 660 24 Periodic PCRS/STA alignment filter updatesPCS-105 12 70 IOC Darks, 70um MIPS-911 30 82 70um IC - routine, Phot AOR MIPS-918 21 112 70um IC - routine, Scan AOR MIPS-915 13 133 70um NF Illumination correction MIPS-313 19 146 70um WF Focal plane survey (coarse) MIPS-132 126 165 Ge Anneal and recovery MIPS-930 12 291 24um Optimum operating temperature MIPS-345 90 10 70um NF Focal plane survey (coarse) MIPS-137 130 100 SED Slit location and orientation MIPS-139 108 230 Periodic PCRS/STA alignment filter updatesPCS-105 12 338 Periodic IRU Calibration updates PCS-055 36 350 IOC Darks, 24um MIPS-910 15 386 24um Flat - routine, Phot AOR MIPS-917 10 401 24um Flat - routine, Scan AOR MIPS-914 13 411 Shutdown activities, Ge and Si MIPS-992 19 Comments: The Ge bias optimization is the dominant task in this campaign. I made a mistake when I wrote it by not performing a thermal anneal at the start of the task. This is corrected by scheduling this task first in the campaign, after the thermal anneals in MIPS-905. The nominal place for the first PCS-105 execution is after MIPS-905, but I prefer to see it before we start so that it does not interfere with our timeline. The second execution of PCS-105 has to happen right after MIPS-246. There is a 70um thermal anneal 70m before the end of MIPS-246 and a 160um anneal 10m before the end of the task. With about 9h of Ge activities in the rest of the campaign, it is essential that we take advantage of these anneals and defer all 24um tasks until the end of the campaign. Unfortunately the 24um optimum temperature test is primarily a 160um dark measurement test, so it has to come early after an anneal. I start with the 70um darks because we were pointing at dark sky for MIPS-246. I placed the three ICs next, grouped to reduce pointing overheads. The WF FPS comes next. This is over 13h since the start of the campaign, so operating temperatures should be stabilized. An anneal is needed next, followed by the 24um optimum operating temperature test. I put this test right after an anneal to obtain the best possible information on the 160um dark levels and to avoid having to perform an extra thermal anneal. The 70um NF FPS task comes next. I am eager to complete this task before too much radiation damage has accumulated. The SED slit location is the last of the Ge tasks. This test runs from 4-5.5h after the thermal anneal. I think we can get away with this even if the 70um array proves to be rather sensitive to radiation damage because the analysis will involve averaging over a large number of pixels. That makes us relatively immune to spiking events, which is not the case for the FPS. This brings us to 9h after the previous PCRS calibration so the IRU and PCRS tasks have to come next. I listed the PCRS task first since it is supposed to happen on an 8h schedule. For the Si tasks, the flats should be paired. The darks were listed first because both the darks and the SED task are likely to be performed near the CVZ, however the PCS tasks involve pointing elsewhere in the sky, so we could just as well perform the flats first before the darks. --------------------------------------------------------------------- MIPS CAMPAIGN J Time since Start-up activities, Si and Ge MIPS-905 35 Ge anneal IOC Darks, 160um MIPS-912 15 24 IOC Darks, 70um MIPS-911 30 39 IOC Darks, 24um MIPS-910 15 69 24um Flat - routine, Scan AOR MIPS-914 13 84 24um Flat - routine, Phot AOR MIPS-917 10 97 160um IC - routine, Scan AOR MIPS-916 13 107 160um IC - routine, Phot AOR MIPS-919 21 120 70um IC - routine, Scan AOR MIPS-915 13 141 70um IC - routine, Phot AOR MIPS-918 21 154 70um NF Illumination correction MIPS-313 19 165 Periodic PCRS/STA alignment filter updatesPCS-105 12 184 Ge Anneal and recovery MIPS-930 12 196 160um Flux standards MIPS-924 20 10 IER to set 160um stim level to 0x2E 0 30 160um Stim and star calibration MIPS-601 7 30 IER to set 160um stim level to 0x33 0 37 160um Stim and star calibration MIPS-601 7 37 IER to set 160um stim level to 0x3C 0 44 160um Stim and star calibration MIPS-601 7 44 IER to reset 160um stim level to 0x37 0 51 70um Flux standards MIPS-922 12 51 IER to set 70um stim level to 0x22 0 63 70um Stim and star calibration MIPS-600 10 63 IER to set 70um stim level to 0x2A 0 73 70um Stim and star calibration MIPS-600 10 73 IER to set 70um stim level to 0x34 0 83 70um Stim and star calibration MIPS-600 10 83 IER to reset 70um stim level to 0x2F 0 93 Routine 24um Flux standard MIPS-920 11 93 70um NF Focal plane survey (coarse) MIPS-137 130 104 Ge Anneal and recovery MIPS-930 12 234 160um Array location and orientation MIPS-234 152 10 SED Slit location and orientation MIPS-139 108 162 70um WF Frame table update sanity check MIPS-961 8 270 Periodic PCRS/STA alignment filter updatesPCS-105 12 278 Periodic IRU Calibration updates PCS-055 36 290 Ge Anneal and recovery MIPS-930 12 326 160um Flux standards MIPS-924 20 10 IER to set 160um stim level to 0x2E 0 30 160um Stim and star calibration MIPS-601 7 30 IER to set 160um stim level to 0x33 0 37 160um Stim and star calibration MIPS-601 7 37 IER to set 160um stim level to 0x3C 0 44 160um Stim and star calibration MIPS-601 7 44 IER to reset 160um stim level to 0x37 0 51 70um Flux standards MIPS-922 12 51 IER to set 70um stim level to 0x22 0 63 70um Stim and star calibration MIPS-600 10 63 IER to set 70um stim level to 0x2A 0 73 70um Stim and star calibration MIPS-600 10 73 IER to set 70um stim level to 0x34 0 83 70um Stim and star calibration MIPS-600 10 83 IER to reset 70um stim level to 0x2F 0 93 Shutdown activities, Ge and Si MIPS-992 19 93 Comments: There are two primary activities in this campaign. One is a very thorough investigation of calibration using standard activities, the other is a set of focal plane surveys. For the calibration tasks, I used the standard startup sequence as much as possible. I put the Phot versions of the 24um flat, 70um IC, and 160um IC in sequence right after the Scan versions of each task. This maximizes efficiency and provides the best comparison between the two techniques. The extra time required is not that large, so the timing of the calibration activities is not altered much compared to the standard startup. After the first anneal, I broke from the usual startup sequence by running the set of MIPS-601 tasks immediately after MIPS-924. I jumped from there to MIPS-922, postponing MIPS-920 until later. This puts MIPS-922 at very nearly the usual timing after the thermal anneal. Both of these Ge calibration sequences are completed before much cosmic ray damage has accumulated. The next task is the 24um flux standard. This is very different timing from the usual startup sequence, but that should not be an issue at 24um this long after the initial powerup. The 70um NF FPS was placed next, partly because the pointing is likely to be similar to that of the flux standards but mostly because it fit nicely into the available time. This task is run between 2-4h after the thermal anneal, which is very good timing. The task starts about 5h after startup, which ought to be sufficient to reach temperature stability. A thermal anneal prepares the 160um array for MIPS-234. This task is long enough that it really needed to be run right after a thermal anneal. The SED slit location task will have similar pointing and will work well between 3-4.5h after the anneal. The frame table update comes next and was placed here to avoid interfering with any of the other calibration activities. The campaign ends with a repeat of the Ge flux standard calibration activities, starting with a thermal anneal and following the same sequence as was used earlier. This cycle of calibration tasks starts about 7.5h after the start of the first set, so we have a decent timeline for looking for changes. --------------------------------------------------------------------- MIPS CAMPAIGN K Time since Start-up activities, Si and Ge MIPS-905 35 Ge anneal Startup Transients, 24um and Ge Arrays MIPS-226 60 24 24um Flat - routine, Scan AOR MIPS-914 13 84 24um Flat - routine, Phot AOR MIPS-917 10 97 160um IC - routine, Scan AOR MIPS-916 13 107 160um IC - routine, Phot AOR MIPS-919 21 120 70um IC - routine, Scan AOR MIPS-915 13 141 70um IC - routine, Phot AOR MIPS-918 21 154 70um NF Illumination correction MIPS-313 19 175 SED Mode Wavelength Calibration MIPS-145 40 194 Ge Anneal and recovery MIPS-930 12 234 Flux Calibration 1 MIPS-350 83 10 160um Spectral Leak Check MIPS-242 50 93 Periodic PCRS/STA alignment filter updatesPCS-105 12 143 SED Slit Illumination Function MIPS-299 20 155 SED Skeleton AOT Validation MIPS-300 42 175 70um and 160um Thermal Anneal Behavior MIPS-160 360 0 Periodic PCRS/STA alignment filter updatesPCS-105 12 360 Periodic IRU Calibration updates PCS-055 36 372 Ge Anneal and recovery MIPS-930 12 408 IOC Darks, 160um MIPS-912 15 10 160um Array Location and Orientation MIPS-234 152 25 IOC Darks, 70um MIPS-911 30 177 IOC Darks, 24um MIPS-910 15 207 Shutdown activities, Ge and Si MIPS-992 19 222 Comments: A big part of this campaign is checking the stability and behavior of the Ge arrays following startup and following a thermal anneal. MIPS-226 has to come right away to address this need, and MIPS-160 has to come much later in the campaign, when startup transients are not a factor. None of the other tasks impose major scheduling constraints. I changed the usual startup sequence because of the 1-hour hit caused by MIPS-226. By displacing the dark tasks, I maintain the same schedule for the flats and ICs as was used in campaign J. There is no reason to perform an anneal yet, so I placed one of the SED mode tests next. I chose the wavelength calibration task because the PN or HII region in question might be in a similar part of the sky to the IC fields, whereas the other SED tasks are probably in the CVZ. I placed an anneal next rather than the other two SED tasks to even out the time in each anneal cycle. After the anneal, the first task is the flux calibration task MIPS-350. This task is a superset of the usual flux standard tasks MIPS-920,922,924, which are performed right after the anneal in the standard startup sequence. To match the usual sequence and to assure best array performance, it would be best if the MIPS-350 observations were performed in the sequence 160um first, then 24um and 70um. The 160um spectral leak check comes next. It is a 160um task and so needs to be run within the first two hours after an anneal. The other two SED tasks finish up this anneal sequence. This brings us about 7h into the campaign, which is sufficiently long for us to be able to place MIPS-160 next. A lot of the trouble in scheduling this campaign comes from trying to fit in the PCRS/IRU calibrations, with the 6h duration MIPS-160 being the biggest problem. By placing MIPS-160 here, we can get the PCS tasks out of the way before the next thermal anneal. The 160um darks come first after the anneal, just as they do at the start of a normal campaign. The 160um FPS has to come next since it runs 2.5h and needs to be completed within the first 3h after an anneal. It is placed well after the start of the campaign, so there should be very good temperature stability. The other two darks finish off the campaign. --------------------------------------------------------------------- Standard MIPS CAMPAIGN Calibration Activities Time since Start-up activities, Si and Ge MIPS-905 35 Ge anneal - 10m R70 R160 IOC Darks, 160um MIPS-912 15 14 1.06 1.40 IOC Darks, 70um MIPS-911 30 29 1.12 1.80 IOC Darks, 24um MIPS-910 15 59 1.23 2.53 24um Flat - routine, Scan AOR MIPS-914 13 74 1.29 2.86 160um IC - routine, Scan AOR MIPS-916 13 87 1.33 3.13 70um IC - routine, Scan AOR MIPS-915 13 100 1.38 3.39 70um NF Illumination correction MIPS-313 19 113 1.42 3.63 Ge Anneal and recovery MIPS-930 12 132 160um Flux standards MIPS-924 20 0 1.00 1.00 Routine 24um Flux standard MIPS-920 11 20 1.08 1.56 24um Linearity 30 31 1.13 1.85 70um Flux standards MIPS-922 12 61 1.24 2.57 Comments: Our usual Ge anneal activity includes 10m of settling time after the completion of the anneals. For purposes of this discussion, I assume that the responsivity at that time has a value of 1.0 and that it increases as R(t) = R0 + 7R0[1-exp(-t/Tau)], where Tau=240m at 160um and as R(t) = R0 + 2R0[1-exp(-t/Tau)], where Tau=480m at 70um. This is not a great description of the responsivity behavior (it fit very well to the UC Davis 160um data with a time constant of 60m, but the energy deposition and hit rate were each about 2-3x too high there; at 70um only about half of the response change comes from cosmic rays, and the detector relaxation time constant and the cosmic ray time constants are probably different from each other). There is 10m of settling time after the end of the Vrst calibration in the startup activities. Ge darks can be a little squirrely depending on what was observed before them, but in this case we should have been pointing at fairly low brightness sky for the previous 10m, which is probably ideal for settling the detectors. When do we point to the low brightness sky? I am assuming that it will be before the end of MIPS-905, but I really don't know. The darks should be done in sequence to reduce pointing overheads and to minimize latency effects. Cosmic ray effects are minimized by doing the 160um darks first, followed by the 70um and 24um darks. By measuring the darks this soon after the anneal, we get a better measure of the zero-point dark current and are less affected by stray light. In standard MIPS operations, we measure point sources and stim flashes in the Ge detector quick response regime, while sky backgrounds and flatfields/illumination corrections are measured in the slow response regime. Chad Engelbracht did an analysis of Ge data taken with a modest background and with regular stim flashes. He found that the stim flashes did a poor job of tracking the responsivity of the background for a short time after a thermal anneal, but after about 15m at 160um and an hour at 70um the stim flashes were quite effective at calibrating the background. Based on this result, it appears beneficial to wait at least an hour and preferably longer after a thermal anneal before collecting our 70um and 160um IC data. The worry here is that the shape of the measured IC will change during the first hour after an anneal due to pixel-to-pixel differences in the settling time. The IC measured after an hour should be fairly stable and should be good for correcting both point sources and backgrounds. Most of our science data will be taken in this "settled" regime, so it is beneficial to measuring the ICs in this same regime. To accomplish these goals, it is best to delay the IC measurements by measuring the 24um flats first. The 160um settling time is quick, so the 160um IC comes next. The 70um ICs are saved for last. The flux standards come last. I injected an anneal here since otherwise the 160um responsivity would be 4.0 and the 70um responsivity 1.55 during the standard star measurements. Measuring the standards right after an anneal gives us a good zero-point calibration of the stims and responsivity. By placing the 24um standard star measurement this late into the startup calibrations, it is nearly 3h after the instrument startup. Ground-based tests suggest a drop in 24um responsivity by 1% during the first hour of operations followed by recovery over the next two hours. This test would occur when the responsivity is at very close to its final value. The stray light levels and dark current should be low enough during the 24um darks that a 1% drop in responsivity should be difficult to measure. Since the drop in responsivity was an array-wide effect with little pixel-to-pixel scatter, it should have no effect on the 24um flats. For the flux calibrators, there are two important issues. The first is the calibration repeatability of the stim flashes after an anneal, which Chad found to be good immediately after an anneal at 160um and good after about 10m at 70um. The other factor is the stability of the stim/source ratio, which has a settling time akin to the stim/background ratio for any source not equal in brightness to the stim. Neither of these factors precludes running the 160um flux standards first, and that has the advantage of completing the observations before the responsivity has changed much. The 70um stim/source ratio will not be settled yet, so the two 24um tests are run next. The 70um flux standards are measured 71m after the anneal, which is long enough to give good stability. I am not sure what the 24um linearity test entails but it seems possible that we will replace it by including RAW measurements in one of our other 24um tasks. ---------------------------------------------------------------------