Log on to plus1 or plus2 and then to vme1: o There are some environmental variables which need to be set, ROOTSYS and LD_LIBRARY_PATH, which can be found in my .bash_profile. It is also useful to set the PATH to include the executable areas of the online, again done in my .bash_profile. In fact, it is probably easiest to log on as dauncey and get them automatically. o Use >ssh -X in both cases so as to get X windows back for the online monitoring. o Once on vme1 >cd online to get to the DAQ working directory. Start up the DAQ background process: o Type >startUp and wait for ~10 secs: it should give something like >startUp Deleting payloads Log file used: data/log/Log1144228878.out where the first line will only appear if the process was not running previously. This runs a background task called runner which should show up with >ps The std output is sent to the log file named and there is a corresponding .err file for the std error. o A basic check everything is OK is to do >grep Firmware data/log/Log1144228878.out (for the actual log file being used) which should give something like >grep Firmware data/log/Log1144228878.out Firmware id = 553648918 = 0x21000316 = 0b00100001 00000000... Firmware id = 302019397 = 0x12007345 = 0b00010010 00000000... Firmware id = 302579730 = 0x12090012 = 0b00010010 00001001... Firmware date = 01133876981 = Tue Dec 6 13:49:41 2005 Firmware Id = 234840467 = 0x0dff6193 = 0b00001101 11111111... Firmware Id = 234840467 = 0x0dff6193 = 0b00001101 11111111... Firmware Id = 234840467 = 0x0dff6193 = 0b00001101 11111111... Firmware Id = 234840467 = 0x0dff6193 = 0b00001101 11111111... Firmware Id = 234840467 = 0x0dff6193 = 0b00001101 11111111... Firmware Id = 234840467 = 0x0dff6193 = 0b00001101 11111111... Firmware Id = 234840467 = 0x0dff6193 = 0b00001101 11111111... Firmware Id = 234840467 = 0x0dff6193 = 0b00001101 11111111... which are for VME, BE, BETrg (including date) and 8xFEs. o To stop the runner process, then do >shutDown Note, this is done by startUp before restarting runner, so it is not normally required very often. However, do a shutDown before replacing the firmware! o If this does not respond, the emergency stop is to >kill -KILL this process (which won't ever do any damage but is not the clean way to stop!). Starting and stopping a run: o To start a data-taking run, use runStart. For your purposes, the most useful is probably >runStart -w -t crcNoise -v 25 -p 99 where the -w means don't make a raw data output file, -t crcNoise is the run type (here just read the ADCs with no calibration pulses, etc) -v 25 is the version (here bit 0-1 = use the BE software trigger and bit 3-7 = read out all the serial I/O information every N = 1 events) and -p 99 is the print level (99 is maximum, so everything is printed). This is a large amount of printout so the log file will get very big quite quickly and you will only run at around 5Hz. You probably only want to take at most ~1000 events before stopping the run so as to not fill up the disk too much. o The progress of the runner program can be monitored using >runMonitor which simply prints the numbers of each type of record to the screen every five seconds. Ctrl-C to quit out of it. o The basic histograms of the data can be seen doing >hstDisplay HstCrcNoise which should open a ROOT canvas with three plots. The three are the number of times each ADC is read out, the pedestal for that channel and the noise. This updates only every 500 events, so if you are interested in this. it might be better to run with "-v 1" and without the "-p 99" in runStart, which should give a rate around 50Hz. Ctrl-C in the ROOT window or the hstDisplay terminal to stop this. o The run can be stopped using >runEnd This leaves the runner process going, so you can later use runStart again. Looking at the printout: o Use less or emacs or whatever to look at the log file, as given at the startUp time. It will be the last in the data/log/ area, as the large number is the epoch time of the startUp. o There may be several runs in the log file; the easiest way to be sure you are looking at the right one is to search for the string "NEW RUN ==>>" which indicates a new run. Find the one with the right type and start time for the one you are looking for. o To see the configuration data used, then search for the strings "CrcBeConfigurationData" or "CrcFeConfigurationData" and I hope the printout will be relatively easy to interpret. o To jump straight to the first event, search for the string "BE soft trigger for" and immediately after this you will see all the serial I/O data from the VME, BE and FEs, followed a short distance later by the Vlink data. This then repeats for the second event, etc. Note, there will not be BETrg data as this run type uses the BE software trigger. o The standard configuration for the FEs gives 1062 words for the Vlink packet. This is displayed in an interpreted printout to start with and then again as a raw hex dump afterwards.