HyperNews System of H1 The Base release of ReadOut program for FIC8231 for work with H1CDAQ for new "Lumi 2000" subbranch Forum: Detector Components - Luminosity System Date: 2002, Feb 27 From: Alexandre Fomenko Dear All, We have now "the Base release" of ReadOut program for FIC8231 for work with H1CDAQ for new "Lumi 2000" subbranch. "The Base release" words means: a) program exists from 20 November 2001 afternoon - when new subbranch "Lumi 2000" was included into H1CDAQ by Igor Sheviakov after next hardware tuning. - Run 291577 - first H1Run at 2001 where first BOS banks from new Lumi system (only for RunStart and RunEnd records) had appeared through 4th subbranch at 11th H1CDAQ branch. It had happened anywhere 20.11.2001 14:58. DAQSUB = 20400 (17th branch is ON, 9th branch - "old" Lumi - is OFF) ... 4C55504F:LUPO 2 9 9 9 4C555452:LUTR 2 40 40 40 ... Note: at latest releases I had changed names of that BOS banks on LNPO anf LNTR - Run 292058 - first H1Run at 2001 with BOS banks at each H1 event (22 Nov 2001 16:11-16:38) DAQSUB = 20300 XI = 8F2 (17th branch is ON, 10th branch with 9th subbranch - "old" Lumi - is OFF) 4C524545:LREE 7974 13 13 13 4C525045:LRPE 7974 25 25 25 4C55504F:LUPO 1 9 9 9 4C555452:LUTR a1 40 40 40 Note: at latest releses I had changed names of that BOS banks on LNEE,LNPE,LNPO & LNTR Executable is created at aplumi03.desy.de (H1lumi) MacII MPW-environment and can be loaded to FIC8231 and started to run there from MPW-environment of that MacII using Bill's VME_tools. Executable is kept at hard disc of aplumi03.desy.de MacII. b) when somebody at H1 Shift Crew includes "Lumi 2000" into H1CDAQ operation at H1 supervisor control panel: 1. it looks like normal operation of ordinary subbranch 2. subbranch 17 ("Lumi 2000" has that number now) can react on all interrupts (PrepRun, L2Keep, L3Keep, L3Reject, StopRun, AbortRun) 3. VMEtaxi interface is operated with MEBBASE =0xD0180000 (as it seen from FIC8231) inside 11th branch of H1CDAQ (as 4th subbranch). 4. some BOSbanks (LNTR,LNPO) are written into StartRunRecord and into EndRunRecord. 5. On every L2Keep interrupt the program can read data from three neighbour bunch-xings at ADCs pipelines for all 24 PD channels and all 12 ET channels, can decode and process (pedestal subtraction) that data and can create BOS banks - TSTC,LNEE,LNEF,LNPE,LNPF and can send that BOS banks into every event at H1DataFlow through VMEtaxi protocol. c) BUT nothing "interesting" is at all BOS banks which are sent now to H1DataFlow (may be except TSTC BOS bank): 1. Contents of TSTC BOS bank is correct - cause there are no problems with event building at H1 CDAQ when "Lumi 2000" is included into ReadOut. 2. Alan was informed about new names of BOS banks and it's formats (LNTR.LNPO,LNEE,LNPE,LNEF,LNPF) for "no warning messages" during Logging task during H1Data Taking. At future some from that that BOS banks names will be changed on "ordinary"(old) ones for H1REC - LNEE-> LREE and LNPE -> LRPE same as it were used at old Luminosity subbranch - Vladimir had asked to keep old names for LREE and LRPE BOS banks - for his new reconstruction program code. For that the creation of LREE,LRPE for H1DataFlow at "old" Luminosity branch (FNC branch now) should be done. 3. I did "nothing" with preparing of contents for LNTR and LNPO BOS banks - they are existing as templates where later we can put later through VME DualPortedMemory something useful (into StartRun and at EndRun Records) 4. Contents of LNEE,LNPE,LNEF,LNPF BOS banks are pedestal data (at best case - if ADC_Ramp (H1Lumi local data taking) not operating - or something random and not predictable at case ADC_Ramp running (cause reading is made from non-stopped pipelines) 5. Special tuned delay constants should be defined for PD and ET pipelines (how far T0+1,T0,T0-1 data are disposed from Front of data at pipelines - at 48 nsec units - value from 0 to 511) - some tools are prepared for that. Trigger bits from PD or ET should be used as source of L2Keep during that tuning procedure. Just now some 'not correct' values are used (only for test and time stability operation) 6. There are not templates of KEEP BOS banks for "non-event" data still - under dvelopment (both sides - at LynxOS and FIC 8231) - There are proposals of needed KEEP banks for Luminosity data, for Calibration data, for Polarization data, for Timing data etc. (see "HyperNews/Forum: Detector Components - Luminosity System", message from Arnd - "Minutes of the Lumi system meeting on Jan 22, 2002". Some special program (or fragment of some program) at LynxOS station (aplumi02.desy.de) will prepare data with needed update frequency at DualPorted Memory and FIC8231 will retranslate them into H1DataFlow as KEEP BOS banks. I don't know more information as it written there about KEEP BOS bank formats. d) Igor Sheviakov had estimated - 03.12.2001 - at one from H1CentralTrigger application windows - mean readout time for branch "Lumi 2000" is near 818 mksec - from some of H1Runs at that day. e) latest and freshest release of ReadOut Program had worked starting from H1 Run 294918 (07 Dec 2001 11:21) Release is keeping at aplumi03.desy.de Mac II directory: /Desktop/HD_LPI5/MPW/Release_2001/Centraldaq_2001_0612 executable TwinFIC1new.........MPST MPS .....59442 Fri, Dec 07, 11:13 AM 2001 script for run Reboot_FIC1.........TEXT MPS .......655 Tue, Dec 04, 11:42 AM 2001 As example: Important fragments from cartl's RunSummary for H1Run 294885 ("old" Luminosity branch and "new" Lumi 2000" were included into H1CDAQ simultaneously) NRUN NrofEvts Ev.Fst Ev.Lst DATE TIME <---> DATE TIME 294885 794 2 80000000 11206 233855 11206 233927 DAQSUB = 30F00 ... 4C464E43:LFNC 779 584 584 584 4C523145:LR1E 779 8 7 9 4C523150:LR1P 779 5 5 5 4C523246:LR2F 779 63 63 63 4C524545:LREE 779 36 27 42 4C524550:LREP 779 26 26 26 4C525045:LRPE 779 26 21 28 4C525050:LRPP 779 14 14 14 4C525446:LRTF 78 13 13 13 4C52544E:LRTN 78 16 16 16 4C55504F:LUPO 1 9 9 9 4C555452:LUTR 1 40 40 40 56455445:VETE 779 29 29 29 ... 4C4E4545:LNEE 779 13 13 13 <- new 4C4E4546:LNEF 779 85 85 85 <- new 4C4E5045:LNPE 779 25 25 25 <- new 4C4E5046:LNPF 779 169 169 169 <- new 4C4E504F:LNPO 1 9 9 9 <- new 4C4E5452:LNTR 1 40 40 40 <- new .... f) I am using standard Bill's VME-tools for monitoring of FIC8231 program activity now (for debugging purposes). At future - when a lot of everything will be more clear as now - there are ideas to make some monitoring tool at Java environment (server at LynxOS station, client with Java's GUI at any h1wgs) - skeleton of that monitoring tool was developed and tested with simpliest things (High Voltage status monitoring). g) There are plans to introduce "WatchDog"-program at MacII aplumi03.desy.de to provide automatic restart of program after crash - as it was made for old system (similar one as Igor Sheviakov's MPW environment executables "DAQ_Control" with Bill's VME tools using) h) latest release and tools for pipeline delay constant tuning program were tested last time at 28 January 2002 after Winter 2001/2002 ShutDown - everything is operating (still): as example - the set of pedestals for neigbour T0-1,T0,T0+1 bunch x-ings (12 samplings - 4 samplings per each bunch x-ing) from ADC[1],ADC[2], ADC[3].ADC[4] (Photon Detector 24 channels -PDx1-PDx12,PDy1-PDy12) and from ADC[6] ADC[7] (ET6 Detector channels ET1-ET12): These data are retranslated into H1DataFlow through LNPF and LNEF BOS-banks Photon Detector ADCs Data [LNPF bank contents] 24 25 25 24 24 25 27 27 25 23 24 23 52 54 54 53 53 53 55 53 52 52 53 54 41 43 43 42 43 43 45 43 40 41 43 42 71 71 74 75 74 74 74 73 72 72 75 73 70 69 71 71 71 71 73 71 68 70 69 71 49 47 50 49 49 48 50 49 48 46 48 47 17 19 20 17 18 20 20 19 18 17 17 17 52 56 57 55 56 56 58 59 57 53 54 54 109 110 110 110 109 108 109 111 107 108 106 109 74 74 74 76 75 74 76 77 75 72 75 73 39 42 42 41 40 42 42 42 41 39 42 41 67 69 71 71 70 71 70 70 69 66 68 70 127 126 120 122 123 125 121 119 120 123 124 118 59 58 56 60 58 60 54 55 56 60 58 54 181 179 176 177 177 180 176 174 178 179 178 175 65 65 63 62 65 65 63 60 61 63 63 58 143 143 142 141 143 143 142 138 141 143 142 138 159 158 159 158 157 160 158 156 156 159 157 153 189 188 186 187 189 189 184 186 185 187 185 184 163 162 161 162 162 164 161 161 160 163 164 160 152 152 151 152 151 153 152 149 147 149 150 147 141 139 137 137 140 142 136 139 136 139 137 134 151 152 149 149 153 155 151 148 149 151 150 148 103 100 101 102 103 102 100 100 100 101 101 97 ET6 Detector ADCs Data [LNEF BOS ban contents] 134 136 137 137 135 136 134 137 131 132 135 135 159 164 162 161 159 161 158 159 159 160 160 159 154 155 157 154 153 154 155 156 150 153 152 154 142 143 145 142 140 142 143 143 141 140 144 142 127 129 131 130 129 127 128 130 127 126 127 127 120 121 122 121 121 121 122 121 120 119 119 120 103 105 106 104 105 105 105 107 102 102 105 105 173 176 175 172 176 173 175 175 172 173 175 173 116 118 119 114 114 116 117 119 116 114 116 117 120 122 124 119 122 121 120 120 119 120 118 121 87 90 93 89 92 89 89 90 88 87 91 91 130 132 132 132 132 132 131 132 131 128 132 130 i) "base program" is still under development now - every time when I want to test new release - I should inform H1ShiftCrew and to ask to exclude Lumi 2000 from H1 ReadOut on some period or to include it again. j) As conclusion - in principle a lot of everything is working but a lot of everything should be added or developed or changed etc. Sincerely, Alexander Fomenko [Back to overview of forums] to: "The Base release of ReadOut program for FIC8231 for work with H1CDAQ for new "Lumi 2000" subbranch" Thomas Kluge