Since the CW viewer camera looking at the C dump has failed, I looked at the horizontal signals from the last three BPMs in the line, 3H07C (standard), 3H08 (big), 3H09. I think the latter two overstate position by a factor of two or so but http://opsweb.acc.jlab.org/CSUEApps/atlis/task/17928 will have to be run to determine that.
Figure 1 shows one week, periodicity in 3H08 (small) and 3H09 (large). Figure 2 shows one day. In figure 3 I added the other four SHMS magnets to figure 2. No periodicity at this coarse current scale.
Figure 1, top plot, shows total CEBAF current in blue (0R08), Hall A current in gold and Hall C in green, all on the same axis. The bottom plot in pink shows BLM 3C12A trips, green is NL RF, purple SL RF, and blue-gray Inj RF. A spike in pink without an associated spike in the other three indicates a 3C12A BLM beam trip alone, no accompanying RF cuplrit.
Around 0400, when only 52uA of beam were in the machine, we saw next to no 3C12A-exclusive beam trips. More than this, however, and the frequency of 3C12A-exclusive trips goes up dramatically, like between 0700 and 0800.
Figures 1 and 2 show the fault rates I model. Ops has been turning down 2L25-1, the biggest offender. 2L20 is the lock zone; it's been pushed up. Figure 3 shows lem predictions including many small ODVH changes I just made to redistribute gradient upon next lem. Figure 4 shows that dropping the current used by lem from 600 uA to 540 uA, still 50% above the 360 uA we've been running in the linacs, would decrease trips one/hour. If we can get up to 410 uA in linac, 82*5, we'd still have ~30% margin for input Q, gradient calibration and tube power errors.
THOMAS JEFFERSON NATIONAL ACCELERATOR FACILITY
ACCELERATOR OPERATIONS CREW CHIEF SIGN IN FORM
Date : 02-18-18
Time : 07:24:28
Day : Sunday
Shift : Day
Crew Chief : Carlino
Program Deputy: Bickley
Team : X
Arm : Anthony
LERF Operator: NA
Operators : Anthony
This stamp certifies that the above Crew Chief has read/reviewed/received the following: