HADES Test beamtime - DCS lessons learned / truisms "Binsenweisheiten" / anecdotes

• Peter Zumbruch (GSI Helmholtzzentrum für Schwerionenforschung GmbH(GSI))

Room: ZOOM HADES just finished its test beamtime (Feb 8-13, 2021) introducing and evaluating the performance of the new forward detectors together at maximum proton beam intensities 10⁸ and maximum beam energies of 4.2 GeV on an LH2 target. ## beam vs. equipment - again, while threading the beam to the target, powered LV devices (TDK lamdda) where hit by intense primary/secondary beam 2 out of 8 failed (positioned there for more then 15 years) - "again", since Mar2019, magnet failure in front of the cave, opened the way directly to a stack of ~10 TDK Lambda, ~4-5 broken, power amplifier. ### Options → evaluation of different positioning longer cables / rules for switching power → shielding → "TOP" principle (from OSH/"Arbeitsschutz"):Technical → Organizational (→ Personal)) ## Silence - 1-wire sensors (performance & keep alive) - based on HadCon2: chained 1-wire bus readout of ADCs - EPICS uses streamDevice to communicate with HadCon2 - record issues read command, every second. - HadCon2 (1-wire master) evaluates itself list of available devices and addresses and reads the values and publishes the results "en bloc". - StreamDevice triggers I/O Events to scan those results and processes only the relevant records. ### → Problem: no error recognition - If a sensor or the complete bus does not answer anymore, it is not recognized immediately, since it is not updated. ### → possible solutions: - single device read/response mechanism vs. performance? - regular separate device check, changing the error state of the value records. ## remotely switchable plugs - Latter problem was temporarily solved by power cycle to revive the HadCon2 connection ### → switchable plug plug + automatic IOC restart ## (this time no network issues)

a.o.b Room: ZOOM