File : cal_EA-T9_9611.txt ------------------------------------------------------ Testbeam November, 1996 in East Area, T9 from : November, 18. 96 to : November, 27. 96 ------------------------------------------------------ Tracker, X-Pixel pixel on growth side, 150 micron pitch ds=? on VA3 in Scirocco 8 : (121+-10) e/ADC (TP,961218) Tracker, NDF005-F: strips on growth side ds=390 micron, CrAu, pitch 50 micron on VA2 : (148+-15) e/ADC (TP,961128) Tracker, DBDS-39: strips on growth side ds=1180 micron, CrAu, pitch 50 micron on VA2 : (114+-10) e/ADC (TP,961128) info(961204): the calibration value (e/ADC) is different than in last testbeams, since we have new scirocco ADCs. The ADCs now have 12bit (before they had 10 bit). Therefore we have factor 4 higher calibration constants. info(961218): DBDS-39 tracker was on Scirocco 5 F-tracker was on Scirocco 6 X-pixel was on Scirocco 8 Scirocco 5 and Scirocco 6 have the same gain. Scirocco 8 has a different gain than Scirocco 5 (there is a 430 Ohm resistor soldered in Scirocco 8, no circuitry available!) dme/961204 ---------------------------------------------------------------------------------------- From renato@lepsi.in2p3.fr Wed Dec 4 09:52:38 1996 From: Renato Turchetta Subject: telana (fwd) To: Dirk.Meier@cern.ch Date: Wed, 4 Dec 96 9:52:30 MET Mailer: Elm [revision: 70.85] > > Dear Dirk, > sorry for being late. Yesterday we started the analysis. That means the > modifications to TELANA have been done and everything seems to work properly. > The modifications are almost exclusively in the files > TELANA.COMMON > TELANA.DAT > EXA.FOR > with some minor modifications elsewhere. > All the modifications come from the fact that the hardware structure of Siroccos > has been changed. Before each Siroccos had only one channel (i.e. one cable > entereing the Siroccos) and data were coded over 10 bits, resulting in data > stored as 16-bit integer word. > Now there are two kind of Siroccos. > -Group1: One Siroccos has 4 channels and data are converted over 8 bits. The > storage is however done over 32 bits with each word being composed by one 8-bit > word per channel according to the following scheme > > > MSB --------------------------------------------------> LSB > | Channel 3 | Channel 2 | Channel 1 | Channel 0 | > > -Group2: One Siroccos has 2 channels and data are converted over 12 bits. The > storage is however done over 32 bits with each word being composed by one 12-bit > word per channel according to the following scheme > > > MSB --------------------------------------------------> LSB > | Channel 1 MSB | Channel 1 LSB | Channel 0 MSB | Channel 0 LSB | > > > > In the setup used for these runs, there was 1 Sirocco from the group 1 (4 > channel and a maximum of 768 strips per channel) and 3 Siroccos from the group 2 > (2 channels and a amaximum of 256 strips). That is there were 4 Siroccos but 10 > channels (and as usual each channel could be connected to more than one > detector). > > > > Now a channel of Sirocco does not correspond to one Sirocco. Before there were > in the analysis some parameters that were exactly the some: NTOTSIR (the same as > in the data structure) and NSIR, that was only used for the analysis and in > particular was one of the dimensions of the arrays were tha raw data, pedestal, > noise ,.... were put. Since NTOTSIR is equal 4 in the data acquisition and must > be set to 4 in the analysis when reading the data, I decided to give NSIR the > value of 10. In this way I had just to add a few lines in the routine GET_EVENT, > in order to be able to do the correct decoding of the data from the array where > the data are put at the beginning to the array ISIRDAT which is used for the > analysis. This is basically all I did. There were some minor changes de to the > fact that before NTOTSIR and NSIR were numerically the same and then they were > sometimes confused (that is ,e.g., the program was using NTOTSIR even if it was > working on data already passed to the ISIRDAT array). So I add to correct this, > making the program somewhat more logical. A second minor but important > modifications is due to the fact that before data were always on 10 bits with an > 11th bit of overflow. So there were a few checks on the data to see if they were > in this range. Since now data are on 8 or on 12 bits, this check is no longer > correct neither needed. So I took it off . Moreover I had to change the range > for histograms on pedestals which were limited to 1024. > > For the DSTANA program ( or any program you used for the analysis of thhe > DST's), I just changed the maximum number of Siroccos (NSIR variable: it was set > to 8 before and now I raised it to 10). I have not done many checks on this > program but at first sight there is no other problem and analysis shoulsd go > smooth. > > I put a copy of a correctec VMS version of the program under > > CRNVAX::DELPHIH:[SCRATCH.TURCHETTA.ALICE] > > Normally it is not protected so you should get it without any problem. Otherwise > please let me know. > > > Cheers Renato > > > ______________________________________________________________________________ > Renato TURCHETTA Tel:(33)-03-88.10.68.68 > LEPSI (CNRS/University) Fax:(33)-03-88.28.34.85 > 23, rue du Loess E-mail:renato@lepsi.in2p3.fr > F-67037 Strasbourg Cedex E-mail:renato@vxcern.cern.ch > > -- ______________________________________________________________________________ Renato TURCHETTA Tel:(33)-03-88.10.68.68 LEPSI (CNRS/University) Fax:(33)-03-88.28.34.85 23, rue du Loess E-mail:renato@lepsi.in2p3.fr F-67037 Strasbourg Cedex E-mail:renato@vxcern.cern.ch ---------------------------------------------------------------------------------------- From dme@vsnhd4.cern.ch Wed Dec 4 23:57:31 1996 Date: Wed, 4 Dec 96 23:57:01 +0100 From: dme@vsnhd4 (Dirk Meier) To: conway@ruthep.rutgers.edu Cc: kagan@mps.ohio-state.edu, Dirk.Meier@cern.ch Subject: DB39-map, calibration Hi John, here is a map, which could help to identify 'good' regions in the DBDS-39 tracker, good regions are: strip 10..52 strip 54..65 strip 75..86 strip 99..115 strip 119..128 the strip number has an 'unknown' offset relativ to the VA2 channel number. For the calibration I had the following thoughts: The online display of the DAQ is unchanged compared to before we got the new scirocco ADCs. This means there is a range from 0..1023 displayed. Now we have 12 bit ADC, which can cover a range from 0..4095. The least significant bits are cut off by the display (Wojtek told this). Thus the signal S_displayed[ADC] measured in ADC counts, which is displayed by the DAQ, is related to the signal S_reality[ADC] measured in ADC counts, which is the actual signal stored in the file, by 4*S_displayed[ADC]=S_reality[ADC] (1) This is, because the two bits which are cut off increase the available range by a factor of 4. A calibration constant g is given by the signal S[e] measured in electrons and generated via the testpuls, which is related to a signal S[ADC] measured in ADC, counts: g=S[e]/S[ADC] In particular the calibration g_displayed[e/ADC] which I can measure and which I get displayed is given by: g_displayed[e/ADC]=S[e]/S_displayed[ADC] (2) and in particular the calibration g_reality[e/ADC] which we have to use when we look at S_reality[ADC] on the files (like you do right now) is given by : g_reality[e/ADC]=S[e]/S_reality[ADC] (3) When I instert expression (1) into expression (2) and using (3) I get: g_reality[e/ADC]=1/4 * g_displayed[e/ADC] I gave you the calibration g_displayed[e/ADC]=114+-10 e/ADC for DBDS-39. This gives g_reality[e/ADC]=28.5+-2.5 e/ADC. You observe a mean value of the signal distribution {S_reality[ADC]} of =330 ADC. Thus the tracker DBDS-39 gives a mean signal in electrons of = (28.5+-2.5) e/ADC * 330 ADC = 9405 e This is indicates a charge collection distance of 9405/36 mircon = 261 micron. I guess this is not too bad (is 330 ADC really true?). Dirk