Hi, since we are not attending the meeting this afternoon I would like to report in short the status of the luminosity monitor. This is only my personal impression and does not necessarily the represent the opinion of Kams group (so I can't send this to the entire collaboration)... All in all this may seem a bit 'long', and most of this should already be known, but (I guess) was never noted on paper (or a terminal window)... * all three detectors are working fine (or perfect - 'rock steady' one could say...) * calculation of the luminosity from the data file of the histogramming module (program Lum_read_new in lumin/lumscl/check ) - windows are now set for the Psi' energy, but should be readjusted for significant changes in energies (like going from Psi' to chi0) - left and right (fixed) detectors give reliable information on the luminosity (position of the detectors is well known and signal peak is in the area which justifies linear background parametrization) - center (movable) detector: position is not ideal since it put the peak of elastic scattered protons in the area where the background is not linear. Besides that the position of the detector is not known exactly. Therefore the luminosity information of the center detector is not reliable and should not be used until the program is adjusted for these things. At the next possible access Dave will try to move the center detector to a larger angle (relative to 90 degrees) to get the peak in the area with a linear background. Again the position of the detector will only be known through the energy of the peak. He should have contacted one of you for access by now(?). Since the calculated luminosity depends on the correct setting of the windows and parametrization of the background I added a short paw macro to the Lum_read_new packages that allows to look at these 'parameters' and therefore check the reliability of the calculated luminosity values from the program. * 'online' luminosity monitor: Here again the windows used to count the peak have to be adjusted for different energies. In the moment the correct windows for the Psi' and (probably) chi0 are set. The windows for other energies need to be updated. This online luminosity measurement does not make any correction on the background yet. We have a program that should do this correction but it still needs to be tested. * splitter and termination of second output of the splitter: The signal in the first output of the splitter (the output going to the histogramming module) depends on whatever is done with the second output (going to the scalers/mca). Changes in the second output may effect the signal range of the output to the histogramming module. It would be nice to have more control over the splitting but since nobody seems to know anything about the splitter the best thing to do is just to leave everything as it is and not to touch it (or even look at it...) I am looking at the output files from the histogramming module and checking for any systematic effects (not that I expect to find any but if you don't look you will never know...) Just one more point - the luminosity monitor is run by Kam and Dave. All information I get (usually) comes from Dave and from being 'on site'. I usually have no insight on Kams plans however. - Have a nice day, Willi --------------------------------------------------------------------------- here have been two problems in the CCAL during this stack which may affect any quick turn-around analyses being performed. - First, a shaper board with the wrong RC values was installed into the system before the run. (affects runs 3235-3245) It has been replaced with a spare after the second deceleration. (16 blocks) - Second, 4 blocks (from the same HV module) had a voltage readback 60 V higher than the set value. Since the gain constants currently in the database do not reflect these large deviations, I would recommend setting the gain of these channels to 0 for the entire stack for the time being. I have prepared a modified offline routine; it is on fn835i: ~michthom/offline/dbccal_5235.FOR. Including this routine in the code when you compile will set the gain constants to 0.0 for these 20 blocks. Michelle ---------------------------------------------------------------------------- Hi, I'll briefly report about the situation of the SCI-FI detector. During the evening shift It was decided to turn off the VLPCs because of extra hits in 3 TDCs. The problem was due to an oscillation generated randomly in one of the cassettes (#11). When I came over this morning the oscillation was disappeared (the signal at the scope was nice). The last part of run 5258 was taken with the fibers on and the DAQ didn't complain at all. Other investigation with the pulser didn't show any oscillation. The problem was probably due to an external device (like an RF) that induced that oscillation in the VLPC. ciao Wander ----------------------------------------------------------------------------- 835 Weekly Thursday meeting: 3/16/00 Present: Borreani, Seo, Boca, Kasper, Stancari (M), Hu, Rusack, Vidnovic, Luppi, Negrini, Legger, Baldini, Leger, Gollwitzer, Pordes. Please send corrections/additions to Stephen. Mainly a discussion of previous stack running but first this news given by Keith. No beam for about a week. The "long tanks" in AP-60 have a leak that is believed responsible for the poor beam lifetime (35hours). There will be a search for other leaks in AP-60 in case there are any. Removing, repairing, reinstalling and baking is expected to take a week. The tanks were installed last April. The vacuum has not been good in AP-60 for a while - the present situation is a deterioration of the vacuum to a few 10^-8. Vacuum people will be working up to 12 hour days on the first parts of the process. There will (probably) be supervised access for the next few days but this isn't guaranteed. Seon-Hee asked if this lost time would be "reimbursed". The spokespeople will take this up with the lab at an appropriate time. All Stephen would say is that our best position is to come back with the apparatus in perfect condition and able to take data at the highest luminosity. Discussion of Past Run: ( the run from hell or a good learning experience) Gas-jet: Gabriele: he has installed a new path with the temperature 2 deg. K above the transition curve. The unclogging procedure has been written up - it needs to be performed when intensity drops >10% typically. Takes a few minutes every few hours. Gabriele also said that he had achieved a peak density of >3*10^14 by setting a density of 2.5*10^14 on the slider and then manually lowering the temperature to get nearer (~0.3 deg K above) the saturation curve. Hodoscopes and Cerenkov - Giovanni B. - fine. The Freon consumption seems under control (within a 10% uncertainty). Straws - Federica and Giovanni B. The straws seemed to be working quite well. The HV was raised to improve the efficiency but no numbers were given yet; the occupancies are OK. There was a request to see if the max number of hits to be read-out could be reduced and some discussion on the usefulness of having narrow discriminator pulses. There was some question on just what the HV supply does when the current exceeds the trip limit. There was also the suggestion that the audio alarm be hooked up. Sci-Fi. - Wander. These had been turned off (ie the amplifiers) and removed from the readout because of a major problem seen in the DAQ when the output from one cassette went into continuous mode (oscillation?). This behavior would come and go every few seconds seemingly randomly. It had disappeared when Wander looked at things this morning two hours before the run ended. Wander will inspect the cassettes downstairs to see if there is anything obvious. (see Wander's e-mail) CCAL - Michelle. The gains of the CCAL have been changed to 2.5MeV/ADC count. They had been around 2.8 on average. This is believed responsible for the change (factor of 2) in the ETOT rates (more later). There were some counters whose HV did not come to the correct value - Michelle has a (software) fix for this. (see her e-mail). There is one slot of one shaper crate where the power does not get reliably to the shaper board. A fix needs to be implemented for this. FCAL 2000 - Roger. There were 6 missing channels during the run. Three of these have been fixed (2 were bases that seemed to have been knocked out of position, 1 was a missing connection on a shaper board output paddle). The three remaining missing channels seem to be dead PMT's and are presumably not recoverable. Stephen asked if the situation of the amount of laser light had been worked on and was told no. He suggested that this was an opportunity to do so. Michelle said that there was laser data from the run we just took. Trigger - Wander. The charged trigger seems fine. A problem with the phi-phi trigger may have been a cable which was found to be "inverted". Wander would like to check this with beam, of course. Michelle showed some plots of the timing of the charged signal (the veto) wrt the etot signal. The issue is the loss of neutral events due to accidentals (ie random events) setting the charged veto. In 835 run 1, the loss at 2.5*10^31 was about 12%. By looking at events where the ETOT was set but not the neutral trigger bit (ie events which would have been vetoed by a charged signal) one could see the timing of the charged signal wrt to the trigger. The TDC distribution of the charged veto signal (from H1.H2') has a width (at the base) of ~20 ns wrt the trigger. If she now takes events which satisfied the neutral trigger and looks at when there is a charged signal, the effective veto time - as shown by the period around a neutral trigger when there is no charged signal - is ~40ns (43ns to be precise). Michelle believes this extra width is because the veto width was set by looking at the minbias (ie the trigger strobe) and the hodoscope signals - without any ETOT or other high energy requirement. As such one sees and suffers from all the slewing in the minbias timing near threshold. Looking at the minbias and the hodoscopes triggered on ETOT might give a more appropriate indication of how to set the veto width. This (looking) will be done next stack. This presentation gave rise to a large amount of discussion and Stephen offered another quarter to anyone who could explain some aspects of the plot of charged hit timings on the neutral trigger events - in particular, why the number of hits immediately after the trigger is so much higher than immediately before the trigger and why the distribution after the trigger falls after about 20 ns to a level closer to the before the trigger level. Neutral Trigger: Jason made a remark that he was not able to verify the value of the ETOT high threshold from the combined ETOT lo and hi data. Stephen (and maybe Keith) does not understand this and asked Jason if the recent change in CCAL gains was seen in the turn on curves - of anything. Jason will look to see if such a change can be seen in the superblock thresholds, the ETOT spectrum and possibly elsewhere. He counselled that previous history suggested that run to run changes of 10% occurred without any global gain changes. The issue here is that a) we have to know we are efficient at some reasonable fraction of ETOT and b) given that the ETOT rate has doubled, presumably due to the change in gain, maybe we need to look to see if we should raise the discriminator settings to avoid writing twice as much data as needed. DAQ - Keith. There is probably a bad TDC in the crate readout by DYC15. Finding this may be a long process and will require signals from the Sci-Fi. Wander can arrange for this. There is also probably a bad ADC in another crate (whose number I don't have but Keith knows). Offline - Eleonora, Matteo. It is fine. It was noted that an informational warning about some overflow appeared for the first time after completing the minidst of the last run. Federica said that after relinking the program, this warning did not recur. The yields at each point were not available pending running the data with Michelle's mod for the blocks whose gains were way off due to wrong HV. Pbar Beam - Martin in response to some questions. Martin suggested that it wasn't the jet density as Stephen had told everybody but a number of operational problems that made us miss the energies we planned to sit at. With any luck, these problems should not recur. Giovanni pointed out that missing some points probably allowed us to get data on both sides of the psi' peak. Luminosity Monitor - Keith said that there was a decent luminosity measure on line. He may have suggested that there is a discrepancy between what one gets from the online and the offline. Stephen