Overview | Releases | Download | Docs | Links | Help | RecentChanges

TestingSctRodDaq 4 3 RC2

See also SctRodDaqRelease4.3 and TestingSctRodDaq 4 3 RC1.

TESTING

Installed source release at SR1 (Dave):

Setup file at /daqsoft/sct/setup_RC43.sh

SCT_DAQ_ROOT at /daqsoft/sct/Release43/SctRodDaq

Compilation:

Note, in setup file, I used the -dbg option:
export CMTCONFIG=i686-slc3-gcc323-dbg

make -C $SCT_DAQ_ROOT VmeInterface
make -C $SCT_DAQ_ROOT RodDaq
make -C $SCT_DAQ_ROOT RodUtils
cd $SCT_DAQ_ROOT
make -C build
Built without errors :-)

Start testing:

Use 7 modules as defined in H8Testbox.xml.
Most modules not returning events.... 
Reseat Fibres in BOC - audible click from TX fibre.
Now all 6.5 modules (13 links) returning events.

Run through tests:
NMask ok
PipeLine ok
StrobeDelay ok
3PtGain ok, though high noise modules, many defects.
Occupancy completed, though 'Scan aborted (histo stalled)' mrs at end of test
and data looks rubbish (no threshold scan data below -20mV, see picture below)
SynchTrigger noise - run without errors, but empty data (all histos empty)
DoubleTriggerNoise - ran, but 'errors in event from fibre' for 11 links... one module in CLK/2 mode.
'Probe&Fix' modules from gui, all now returning events, so try again:
DoubleTriggerNoise - 'errors in event from fibre' for 11 links... but all modules are still returning events ok in module probe.

A general complaint - the test result says 'pass' for DoubleTriggerNoise and SynchTriggerNoise even though the data is missing or garbage.

Error during occupancy test (showing 'typical' plot):

http://www.hep.phy.cam.ac.uk/~robinson/snapshot3.png

Standard mrs for occupancy test, but I can't see any pointer to the origin of the error, either here or in the ScanErrors?.txt file (I probably need training at error spotting):

   
ROD (0, 0, 0):  Event 24464    Bin 8
You are about to get the 'Scan aborted (histogramming stalled)' message.  Histogram polling timeout being reported in [/daqsoft/sct/Release43/SctRodDaq/SctApi/SctApiHisto.cxx] at line [2226]. This means that SctApi got bored waiting more than about 5 (search for the second argument of pollHistogramming in SctApiHisto.cxx) seconds for the event counter to be incremented in the rod. Was expecting: finalBin=20, finalTrigger=1000.  Problem was perhaps localised to partition 0 crate0 rod0 run1494 scan7 bin8 event24464
Event errors reported at time: Mon Dec 19 14:27:44 2005
 to /tmp/SctApi.ScanErrors.txt