*** PLEASE READ THE EMAIL ASSOCIATED WITH THE P4 FLARE LINK FOR AN UPDATE TO THE P4 THRESHOLD. *** -------------------------------------------------------------------- To: rac, emartin, andrew, sisafe, ctitask cc: ybutt, plucinsk, buehler@space.mit.edu, pgf@space.mit.edu, das, mwb@space.mit.edu, goeke@space.mit.edu, jimf@space.mit.edu, fergason@space.mit.edu, nousek@astro.psu.edu, eab@space.mit.edu, townsley@astro.psu.edu, dwicker, garmire@astro.psu.edu, wfm@space.mit.edu, broos@astro.psu.edu, da@space.mit.edu, fkb@space.mit.edu, svirani, gchartas@astro.psu.edu, neil.w.tice@lmco.com, larry.g.campbell@lmco.com, isobe, slb, burrows@astro.psu.edu, tsk@astro.psu.edu, mjp@space.mit.edu, gea@space.mit.edu, brm, nss@space.mit.edu Subject: EPHIN thresholds for ACIS safing Date: Wed, 27 Oct 1999 16:58:14 -0400 From: "Paul P. Plucinsky" hi Rob, Shanil has examined the EPHIN data throughout the life of the mission to date with two objectives: 1) to compare the scheduled SI shutdown times derived from the OFLS implementation of the AE8 and AP8 models to the observed count rates from EPHIN 2) to determine thresholds in the SCE1300 and SCP4 channels which would trigger an SI safing event if we entered the belts and the instruments had not be safed for whatever reason, but would NOT have triggered a safemode during the first 70 days of the mission given the observed EPHIN data. Of course, we could use these thresholds to safe the SIs in the event that the particle flux outside the belts was much higher in the future. Unfortunately, Shanil has found that the scheduled shutdown times do not match the observed EPHIN count rates, many times it appears we are turning the SIs on too soon. For example, look at Shanil's www site with these plots, "http://asc.harvard.edu/acis/radbelt/radhist-log.ps" (it is large file). Shanil has plotted horizontal bars indicating when the OFLS schedules the E1,E2,P1 and P2 shutdowns. The OFLS takes the most conservative value and shuts the instruments down for the longest time. Ignore the data before day 228, there is still a problem with the time of the scheduled shutdowns, but the data after day 228 show the effect clearly. As you can see, there are many belt passages in which the instruments should have been off for a longer time. This complicates objective #2 mentioned above, because what do we take as the beginning and end of the "belts", the time RADMON is disabled or the times the EPHIN rates are below some thresholds ? For this study we have adopted the time RADMON is disabled as defining the "belt passage" and Shanil has tabulated the number of times we would have exceeded various thresholds on THREE successive readings (10,10^2,10^3,10^4 counts/s/cm^2/sr ) when Chandra is "outside of the belts (meaning RADMON is enabled)". The results are for SCE1300 and SCP4, starting from Day 228: SCE1300 VIOLATIONS OUTSIDE BELT 10 | 10^2 | 10^3 | 10^4 (counts/s/cm^2/sr) ----------------------- 4 | 2 | 0 | 0 SCP4 10 | 10^2 | 10^3 | 10^4 (counts/s/cm^2/sr) ----------------------- 18 | 14 | 8 | 2 All of these threshold crossings occur very near the belts (either just going in or just coming out) and none of them occur far from the belts. The right answer here is to have a more accurate predictor of the belts which the OFLS will use for scheduling. The SCP41GM will be maintained as a trigger for a hard proton flare and we have not seen anything approaching that yet in the mission, but we are warned to expect such an event during the largest solar flares. As I see it, we have to adopt two plans of attack, one for the near-term to get the safing actions implemented quickly and another for the long term. Remember, that what is important for protecting the CCDs from further damage is that the SIM is translated to the HRC-S position, not that they are powered down. Short-term solution: - move the SIM to HRC-S position 8ks before and after belt passages and conduct ACIS cti measurements (this provides a buffer for inaccurate belt entries and exits) - RADMON should be DISABLED during these time intervals, is this even possible ???? - IF RADMON can be disabled for the cti measurements, set the E1300 threshold at 10 and the SCP4 at 10^3. These levels would have produced NO SI safing actions so far in the mission. [Remember the 4 excursions above 10 in the SCE1300 channel and the 8 excursions above 10^3 in the SCP4 channel all occurred within 8ks of RADMON being enabled upon exiting the "scheduled belts".] - IF RADMON cannot be disabled for the cti measurements, set the E1300 threshold at 10^3 and the SCP4 threshold at 10^5 (We will have to verify with the ACIS team that these levels are acceptable from a CCD damage perspective) Long-term solution: - understand why the OFLS implementation of the AE8 and AP8 models are inaccurate. Yousaf has some insight into the problems with the implementation. - correct the implementation to shut the instruments down when the particle flux is truly high. I suspect the long-term solution will take some time. So, Rob we have two questions for you: 1) Can RADMON be safely disabled during the 8ks cti measurements going into and out of the belts ? 2) We have specified count rates in (counts/s/cm^2/sr), does the OBC use simply rates in counts/integration time ? Does the OBC make geometric factor corrections ? Paul ============================================================================== Paul P. Plucinsky Harvard-Smithsonian Center plucinsky@cfa.harvard.edu for Astrophysics, MS 70 Phone: 617-496-7726 60 Garden Street FAX: 617-495-7356 Cambridge, MA 02138 USA Cellphone: 617-721-4366 Pager: 800-759-8352 PIN 1636493 Pager email: 1636493@skytel.com