WP 6457 ECR E1700032 FRS 7099 DCC T1700025
This model update takes the binary coil status data out of the WD tripping code. If this data goes bad for 60 seconds, a red medm light will show on the ISI platform overview screen labeled OVERTEMP. If the coil driver actually does experience an overtemperature condition, the hardware itself will trip and of course the ISI will trip as the Actuators let go and seismometers rail, you know, cats & dogs living together, mass histeria. So there is little risk that anything too bad would result from this change.
The change was to prevent unnecessary tripping of the ISI when the binary signal from the Coil Driver went erroneously bad. This happened to LHO Oct 2015 & 22977 and the model was changed to allow a 10 sec wait before tripping the watchdog. In May 2016 this was extended to the HAMs. In Sept 2016, this 10 second delay proved not sufficient for LHO BS ISI. It then became a problem for LLO earlier this month. So this is a short likely episode missing history of the problem leading to this model change.
The coilmon status channels are now monitored in DIAG_MAIN, so the control room will get notifications if the status bit goes to zero. We'll need to modify the test code when we do HAMs 2&3.
@SYSDIAG.register_test
def SEI_COILMON_ALL_OK():
"""ISI coilmon status
"""
hams = ['HAM4','HAM5','HAM6']
chambers = ['BS','ITMX','ITMY','ETMX','ETMY']
bscs = chambers
for chamber in bscs:
if ezca['ISI-' + chamber + '_COILMON_STATUS_ALL_OK '] != 1:
yield "ISI %s coilmon drop out" % (chamber)
for chamber in hams:
if ezca['ISI-' +chamber + '_BIO_IN_COILMON_STATUS_ALL_OK'] != 1:
yield "ISI %s coilmon drop out" % (chamber)