Minutes of the Meeting of the Technical Board ============================================= October 31st 2001, 9:30 - 12:15 News from Hera-B (M. Medinnis): ------------------------------- Mike informed us that Peter Buchholz will be the new coordinator. He takes over from K. Ehret who left by end of October. J. Spengler will take care of contact to Hera and act as his deputy. News from Hera (J. Spengler): ----------------------------- Luminosity has been observed both at H1 and Zeus. The specific lumi ( = lumi divided by beam currents) measured is a factor two larger compared to last year, but still a factor two smaller than the expected value. (The goal of the lumi upgrade was to increase luminosity by a factor 4-5). Hera people are fighting with several technical problems like failures of cryo controls, beam losses during injection etc. The Xmas shut-down has to be extended due to important cabling work in the Desy tunnel and in the transfer tunnels affecting the interlock electronics. Consequently, an interlock test of these sections has to take place. Therefore the restart of Hera operation has to be shifted from Jan 7th to Jan 16th 2002. See http://www-hera-b.desy.de/subgroup/running/Run.html. The first target operation is expected for begin of next week. The subgroups are asked to send their requests for the Nov. break (19th - 21st) to the technical coordinator. Target (Yu. Vassiliev): ----------------------- The target mechanics is installed completely and working properly. All counters except 2 of 4 veto counters which give some problems are working under HV and are waiting for stable proton beam for final HV and timing adjustments. The problematic veto counters will be fixed at the next access. However, they do not prevent target operation. The answers to Mike's task list can be found at the "upgrade" web page. The recommissioning of the target operation will start begin of next week. This includes operation of all 8 wires, checks of TaCoS functionality and having a first look at backgrounds, coasting beam and the bunch structure of the proton beam. Vertex (G. Sciacca): ----------------------- A few plots were shown to demonstrate that VDS read-out is working properly with a few minor exceptions which have to be fixed. Two FEDs are missing (config. file has to be updated) and one Sharc board gives problems. The noise distributions of p- and n-side detectors look reasonable, whereas the pedestal distributions of n-sides need further adjustment which will take place during the Nov. break. The VDS task list was presented which can be found in Mike's task list. Inner Tracker (W. Gradl): ------------------------- According to Wolfgang, the assembly of TC stations (MS14+15) is on schedule, MS06 has been installed but not connected and MS05 chambers are equipped with electronics. All stations in the PC area are properly working except MS10 which needs an access. The dry runs have been used to measure the noise of all ITR chambers installed so far. The front-end electronics is very quiet compared to the situation of last year. The width of the raw data distributions is about 4 -5 ADC channels compared to 8 - 14 in 2000. Online sparsification is running (final tuning needed) and CnA to SLT farm is working. The slow control is working reliably in stand-alone mode and will be integrated into common slow control next week. When switching on (reduced) HV, no change in baseline fluc's was observed. Up to now, the low voltage system (very problematic in 2000) gave no problems. See "upgrade" page for the ITR task list. Outer Tracker (H. Kapitza): --------------------------- MC8- is ready for installation. However, all activities concerning MC chambers have been stopped due to the still pending decision about the use of the magnet chambers. The OTR gas system is operated with the nominal gas mixture but with reduced fresh gas addition. The system is handled by our three gas technicians who will take care of all gas systems in the future. HV training of all installed chambers has started about a week ago. They observed essentially no influence of Hera activities except during positron injection. The HV training during the shut-down was stopped in April 2001. 36 groups had been disconnected, 18 of which were turn-on problems. Since then another 6 groups were disconnected during chamber installation. The new HV training gave 6 groups disconnected due to shorts and another 9 "problematic" groups (3 of them are labeled as "hopeless"). All 15 bad groups showed up during the first 2 hours of HV operation. Concerning the read-out, Herbert reported that 70 TDCs in a few crates cannot be read out. The problem is under study (most likely TDC-FCS communication). All ASD-8 boards except MC4-MC8 (see above) were checked. A noise was performed . All PC chambers have a noise level below 1% for a delta threshold of 200 mV. In contrast to this, the situation with the FLT links is worrisome. Despite the modifications there are still a lot of unstable optical links. The commissioning work has to continue till Feb 2002 and needs most likely permanent maintenance. The manpower is hardly sufficient. In order to calibrate and align the OTR chambers, 500 k minimum bias data at 5 MHz are requested. Using these data, a detailed program (hit maps, mask files, t0 calibration, internal alignment, r(t) relation) will be pursued. To establish the operating point requires another 700 k events. Inner high-pt (R. Mizuk): ------------------------- Three of the chambers (6 in total) show large dark currents. PT3- already at reduced HV. The reason for this behavior is not understood because end of August all chambers were OK. Are these HV problems due to humidity. They will modify their gas system to measure the water content of the chamber gas. PT1+ and PT3+ show now a larger noise compared to the situation during shut-down Outer high-pt ( ): ------------------ Alan provided a few transparencies after the meeting. They can be found in the pdf-file containing all presentations. RICH (A. Gorisek): ------------------ Andrej reported on the results of the dry runs. One noisy channel (out of 30k) has been found. By gradually increasing HV they checked for light leaks. Now PM's are operated at nominal high voltage. Taking data, some rings could be seen. A full test of the gas system is planned for end of November. J. Pyrlik and K. Ludwig are preparing this test. Right now, they are discussing how to add the 570 kg C4F10 delivered recently to the gas system which is not completely trivial at low outside temperatures. TRD (V. Savliev): ----------------- Valeri explained that they are going to move their control system over to hall West. There was some delay because their system has been used by Tesla people. ECAL (M. Bruschi, S. Shuvalov): ------------------------------- Marco reported that the commissioning of the ECAL electronics is still not finished. About half of the FED's missing so far have been installed and tested which means that 212 out of 226 are operational. 8 modules came back from CAEN still showing the same problems as before. In total 15 modules are right now at CAEN. Detailed test reports were added in order to make sure that the repair will be successful. Marco expects the ECAL read-out to be complete not before January 2002 which might delay the calibration. 14 out of 128 pretrigger are still missing. Due to lack of manpower no action takes place right now concerning the ECAL inhibit board. In the previous meeting Sergey reported on the large improvements of the ECAL noise. This time he mentioned that this low noise level can be kept even when Hera is operating. However, this picture might change when the final adjustment of timing with respect to BX has been performed. MUON (Y. Zaitsev): ------------------ Yuri showed a list of problems. Some of them will be fixed during the November access, whereas the rest needs more time and has to wait for Xmas. Three pad chambers of MU4 have HV problems, one quadrant of MX4 has to be connected, MU1+ has low voltage problems, MU2- has read-out problems, the gas flow in MU1.4.2 is too low and the shielding wall between MU3 and MU4 does not close properly. As a result of the dry runs, Yuri showed occupancy plots of pad and pixel chambers (LV = 5V, threshold = 1 V). Except the still uncabeled quadrant of MX4 only a few hot spots were found. MUON pretrigger and RICH veto (U. Husemann): -------------------------------------------- The complete pretrigger-FLT connection was tested. A few broken LVDS multiplexers have been shipped to Mannheim for repair. Final commissioning needs target operation in order to finalize lookup tables and channel masks. In parallel to this, the work on the pixel pretrigger is ongoing. The RICH multiplicity veto system is progressing according to schedule. The November access will be used to install the available electronics. Ulrich pointed out that the system can be used not only as veto but as well as an interaction trigger. This system does not depend on the ECAL inhibit board. Development of online software and analysis tolls has started. During 2000 running, two different DMONs and data bases were used to mask hot channels for short (12 min) and long (once per run) time intervals. They plan to unify these different approaches in one system which will be tested by mid November. The old system is still available and was tested during the dry runs. First Level Trigger (B. Schwingenheuer): ---------------------------------------- The analysis of the vector test is almost finished. Some small differences are under study. One of them concerns the double track rejection which depends on the sequence of arrival of the messages. The 2nd TDU was installed and first tests took place. Connection tests are in progress. TC2 cabling was redone which allowed to reduce the latency in case of Muon pretriggers by 4 BX cycles (omit 1 LVDS multiplexer). The upgrade of FLT algorithms is ongoing and they hope to reduce track losses in MU1 and TC2. Second Level Trigger (L. Soezuer): ---------------------------------- After some startup problems, the SLT could be initialized with all CnA tables. The main problems encountered were the 350 SLT histos (very time consuming gatherer subscription) and the much too long transit time from state "INIT" to "CONFIGURE". Plans to fix these problems were presented. They expect to reduce the start-up time by a factor of five. Another project concerns the loading of Muon information into the SLT. It is planned to read-out only selected FEDs according to Muon pretrigger messages. This should raise the rate limit observed in 200 of 10 kHz to about 20 kHz. Fourth Level Trigger (J. Hernandez): ------------------------------------ Jose reported on some work they spend to improve online ARTE. Loading tables from databases is now more stable and when switching off online reconstruction in run control GUI now it's really off. Up to now, it was not switched off completely resulting in a reduced logging time. Reconstruction of MC events is under reconstruction and almost ready. CnA constants are read in using the usual framework. Jose reminded everybody that CnA constants should never be loaded without using the "official" functions. It is forbidden to modify or delete a table in a database. A new version has to be created. Arte online reco consumes > 100 MB. Therefore the Arte memory needs to be monitored because we are not far from running into problems. SLT nodes have 128 MB and L4 nodes 256 MB, but they have to run 2 Arte processes. Everybody is requested to check the memory consumption. Main users are : initialization with GEO ~20 MB, reco ~20 MB, CnA ~30 MB and DQ histos ~5 MB. Data Aquisition (V. Rybnikov): ------------------------------ Vladimir showed a list of the dry runs recorded during the last two weeks (runs 17821 - 17833). One VDS Sharc-board has been exchanged and one OTR board has to be exchanged. There seems to be a shortage of spare Sharc boards. Vladimir and Federico try to recover a few boards by reprogramming them. One board has been send to MSC for repair. The machinery to produce DQ histograms seems to work reliable. As a new feature reference histos can now be displayed. The struggle with our FCS system is still not finished. Rich and high-pt faulty daughters had to be exchanged because there were losses when using pseudo- random triggers. Later on B. Schwingenheuer found out that this problems can be "cured" by reprogramming the mother board. Further tests have to be done. New documentation on data quality (www-hera-b.desy.de/subgroup/DataQuality/ herab/) and a tour on DAQ and slow control (www-hera-b.desy.de/subgroup shiftinfo/sub/DAQ.html are available. J. Spengler November 6th, 2001