| Line 26 was replaced by lines 26-97 |
| - [hrpthmstatus.pdf] |
| + [hrpthmstatus.jpg] |
| + |
| + There is this line reading: Events processed by filler When DAQ is active, the associated number |
| + should increase continuously. Even with SINQ off there should be events any few seconds. If |
| + nothing is coming here, then the electronics does not send neutron events. |
| + |
| + The next line reading Events skipped due to mask is interesting too. This will continuously |
| + increase when the HM sees events where the sync bits are not OK. A small percentage of all neutron |
| + events occuring here is OK. But if all events come here, something is wrong. Scroll down further, and |
| + you can see which sync bit is triggered. But there is only a channel number. What is plugged in |
| + there, ask the electronics guys. |
| + |
| + |
| + |
| + Then there is a line or lines reading something like: |
| + {{{ |
| + bank 0: axis 1: counts low: 0 counts high: 0 |
| + }}} |
| + |
| + Such lines show if there are events outside of the histogramming limits for this axis. |
| + For normal, physical detector axis like x or y, this should be low numbers of neutron events, |
| + ideally 0. If this is not the case, something is wrong: either the electronics delivers phony |
| + numbers of the configuration of the HM is wrong. Check the configuration page. |
| + |
| + For TOF axes it is normal that some percentage of the neutron events are outside of the |
| + time binning range. But if a large number of events is outside, scold the user because she |
| + configured an unsuitable time binning. |
| + |
| + If any of the counters: Events unknown, Packets incomplete or Packets oversized keeps increasing, |
| + then the electronics is sending bad packets on the fibre optics link. After SINQ startup this might |
| + actually have a high count, when there is no electronics switched on or if the fibre optics link |
| + is disconnected this increases too. Due to noise. To be sure, reboot the HM and observe what |
| + happens. |
| + |
| + Then there is the counter with the TSI packets. TSI packets are Time Status Information packets which |
| + are sent by the MDI at regular time intervalls. They contain status information from the electronics. |
| + No matter what, this counter should increase slowly over time. If not, the detector electronics is |
| + either switched off or broken. In the latter case, raise an alarm with electronics guys. This is no longer |
| + a software problem. |
| + |
| + There are other counters here: Detected Fifo Overflows, Detected Taxi Chip Errors. If these |
| + increase, either the fibre optic link has been disconnected or the Mezanine card for reading |
| + the fibre optic link into the HM is broken. Again a problem for the electronics group; not |
| + software. |
| + |
| + !! The Configuration Page |
| + |
| + This page shows the current configuration as known to the HM. It may be useful to look at this |
| + when there is the assumption that the configuration is not properly transferred by SICS. BTW: if there |
| + is such a problem: SICS compiles the XML configuration with a script. The script name can be found by |
| + looking at the initscript parameter of the HM in SICS. |
| + |
| + !! The Debug SHM Tab |
| + |
| + This page shows error messages from the HM software itself. There is rarely something intersting to be |
| + found here: the HM software is stable since years. |
| + |
| + !! The Raw Data Tab |
| + |
| + From this tab, it is possible to look at the raw neutron event messages as they come in from the fibre optic |
| + link. This is sometimes useful for debugging purposes. When you enable Show Packet Info, the packest will |
| + be unpacked into a readable form. Otherwise interpreting the packets is a binary black art |
| + especially as there is not much documentation about the packet format besides the source code of |
| + the SinqHM software. Please note: the WWW-interface works best with firefox; events will only be shown |
| + after restarting DAQ. |
| + |
| + |
| + |
| + |
| + |
| + |
| + |
LIN