RHESSI tohban report, August 19, 2002 Reporting Period: August 11 - 18, 2002 Tohban: Brian Dennis Solar activity: Very low activity initially. Active region 0069 looked promising on Tuesday morning and it produced 2 M-class flares. The Max Millennium forecaster indicated that there was a "chance" of an X flare from this region. A Major Flare Alert was issued on Thursday morning. Many M flares were produced but as of this Monday morning (August 19), the promised X-flare hasn't happened. Observing campaigns: VLA observations by Rob Willson (Tufts University, coordinated with Gordon Holman at Goddard) on August 18, 20, and 22. Spacecraft: nominal Ground station: Berkeley - nominal Wallops - 11-m antenna not available so using the 9-m. Occasional problems with full data recovery. Memory management: The SSR level was reduced to 0% at the end of the last Berkeley pass on Monday, August 12, even with all shutters out. Activity increased dramatically on Tuesday and SSR was close to 50% full on Wednesday morning. This was reduced to 30% by the last pass. It was back up to 64% on Thursday morning and only reduced to 45% at the start of the last pass of the day. The read pointer was moved to a time about 30 minutes before a reported gamma-ray burst to bring the SSR down to the 8% level in anticipation of a major X-class flare. Attenuator operation: All shutters out (attenuator state 0) at start of week. Thin shutter moved in at 08:40 UT on August 15 and operating in the 1-3 mode to keep the memory usage low in anticipation of an X-class flare. Changed to the 0-1-3 mode on Saturday, August 17 at 10:24 UTC. Thin shutter moved out at night-to-day transition at 10:31 UT since SSR was down to 12.5% by the end of the second Wallops pass. Should check on the aspect solution during a big flare to see what the effect of up to 15 shutter motions is on the spin axis. Make sure that we still see 6 limb crossings in the SAS during such a period. Data Gaps: New tohban responsibility to monitor for lost data. Tohban should check for periods for which data are still missing after two days, find out why the data are not in the archive, and determine if and when the missing data will be put in the archive. Any permanent data gaps should be logged in the "gaps.log" file with the reason for the data loss, e.g. the read or write pointer was moved to bring the SSR level down. Not clear yet exactly how to do all this. I reviewed all of the quicklook time curves and identified several data gaps and periods of noisy data each day. Some lined up with Jim's data gaps but many were not included in his list. An examples of noisy data is between 00:25 and 00:35 UT on August 11. A data gap not in Jim's list is at 02:55 to 02:58 UT. I have not tried to track down the causes of these data gaps or why they do not show up in Jim's list. Read pointer was moved from 22:00 UT on August 14 to 10:30 UT on August 15 to provide plenty of memory space in anticipation of an X-class flare. IDPU Reset turned off all events at about 10:05 PST (17:05 UT) on August 15 (see below). All detector HVs turned off as a result of this inadvertent action. Detector HVs were turned on during Berkeley pass on orbit 2865 and the events were enabled during the following Wallops pass at about 10:36 UT on August 16. Other Issues The PMTRAS Counts parameter in the status page often reads zero and shows up as yellow. The correct value of this count is the number of potential stars seen with the PMT RAS in 64 s. It provides the best indication that the PMT RAS is operating correctly. Currently, a special program has to be run manually in order to put this number on the status page. This process should be automated so that the correct number is always indicated, and it should be monitored on a regular basis. This program was run manu ally on Wednesday and now seems to be run routinely. The expected number of star blips is at least one per rotation so the count in 64 s should be at least 18. Suggested Change in Attenuator Control There is a desire to always have the thin shutter out for at least a short time during the daytime part of each orbit. That way, any slowly varying soft X-ray flux can be monitored each orbit even if the attenuators are being operated in the 1-3 mode to conserve SSR space during periods of higher solar activity. This could be achieved by sending timed commands to change the attenuator control state to the 0-1-3 mode sometime during the daytime part of each orbit and then some time later, say 5 minutes, restoring the attenuator control to its previous state. (I understand that an additional command would be required to move the thin attenuator back in again if the previous state was the 1-3 mode.) These commands could be included in the daily command load (that is actually uploaded every 2 days). They would not require any changes to the flight software. SSR Usage During Periods of Low Solar Activity: Discussion of what to do with the available SSR memory resource when the Sun is quiet. Options include the following: 1) Increase SAS rate to 32 readouts per s (units?) from 16 per s to increase accuracy of solar diameter measurements. 2) Stop decimating data during daytime high-latitude parts of the orbit. Both options would require changing the command uploads that are now prepared every two days. Thus, an operational change of this type would take at least two days to implement and two days to reverse unless special procedures were used. Special Activity - IDPU Reset Very low solar activity at the beginning of the week. The thin shutter had been out for the longest time since launch and the SSR was completely emptied at the end of the last pass on Monday. Active region 0069 looked promising on Tuesday morning. In fact, it produced 2 M-class flares and filled the SSR to close to the 50% level by Wednesday morning. This was reduced to 30% by the last Berkeley pass but, perhaps foolishly, the tohban decided to leave the attenuators in the 0-1-3 mode. The Max Millennium forecaster indicated that there was a "chance" of an X flare from this region. The SSR was at 64% at the start of first Berkeley pass on Thursday morning, August 15 at 01:35 am PST. We immediately put the thin attenuator in at 08:40 UT and operated in the 1-3 mode. At the pass starting at 08:13 PST, the SSR was about 45% full. The Max Millennium solar forecaster (Peter Gallagher) had issued a Major Flare Alert stating that there was a good chance of an X-flare from AR 0069 then located at about S07E27. It was decided to move the SSR read pointer to effectively make the full amount of memory available for new data. Jeremy (the operator on duty at the time) pointed out that the HETE home page indicated that there was a "probable" gamma-ray burst recorded at 10:56 UT on August 15. Thus, it was decided to move the read pointer from its then current position at 22:00 UT on August 14 to 10:30 UT on August 15 to be sure that the possible gamma-ray burst data would be saved and telemetered on the following pass. This was done successfully during the Berkeley pass starting at 09:58 am PST, and the SSR level was reduced to 8%. Unfortunately, in attempting to clear the IDPU error counter, the operator (Jeremy) issued the command "IDPU reset" instead of "IDPU clear." Resetting the IDPU puts everything back to the default state. This means that most power to the instrument was turned off. Thus, the detector high voltages were turned off, the SAS, CCD RAS, and the PMT RAS were turned off. Fortunately, the cryocooler power stayed on but the balancer phase and amplitude settings were zeroed out resulting in higher acceleration readings. Mark Lewis and Joe, the other operator quickly figured out what the correct balancer settings were and uploaded them before the end of the pass, thus reducing the acceleration level to the normal value of around 10 mG. Recovery from the IDPU reset took place on the following set of passes starting at 00:04:18 PST. All systems came on and performed nominally. The only anomalies were that the detectors became segmented at lower voltages than previously. Detector 2 became segmented and noisy at its previous voltage and had to be operated at a lower voltage. Detector 5 became noisy at its previous voltage, which was consequently reduced by ~800 volts. Detector resolution was initially reported by David Smith to have been adversely affected in the gamma-ray range (a factor of ~2 degradation) but there appeared to be a full recovery to the original resolution by Saturday, August 17. Tohban next week: TBD