1. Failure analysis at BEPCII
J. Xing, J.S. Cao
IHEP
ARW2017, Versailles,

2. Outline
Introduction
Failure analysis tools
Some examples
Summary

3. Introduction
BEPCII —An upgrade project of Beijing Electron Positron Collider
—A double-ring factory-like machine
—Deliver beams to both HEP & SR
Linac
Storage ring
BESIII
BSRF

4. NEXT : TOP-OFF(Collision)
Design Goals of BEPCII
􀂉 Collision(achieved)
􀂃 Beam energy range GeV(2.3GeV)
􀂃 Optimized beam energy GeV
􀂃 Luminosity ×1033 GeV(April 5, 2016)
􀂃 Full energy injection GeV(2.3GeV)
􀂉 Synchrotron radiation(TOP-UP)
􀂃 Beam energy GeV
􀂃 Beam current mA
􀂃 Keep the existing beam lines unchanged
NEXT : TOP-OFF(Collision) 4

5. Operating Schedules & Management Mechanism
Total 7344hours
( >7000hours/year)
Routine maintenance on Tuesday (68hrs).
Acc.
Daily meeting(Collision).
Weekly meeting(SR).
BEPC National Laboratory (Acc. & HEP & SR & others)
Weekly meeting(Monday) 5

6. Availability of BEPCII(SR)
Average 96%
Goal >98%
Failure rate of BEPCII(Collision)
Average 8%
Goal <5%

7. Failure analysis tools
16-channel oscilloscope
Beam trip diagnostic system 7

8. 16-channel oscilloscope
A 16 channels oscilloscope is used in RF system to record the beam, accelerating voltage, RF power, LLRF control loop related signals and so on;
A preliminary tripping reason analysis can be achieved by distinguishing the time sequences and variation tendencies of the signals. 8

9. Beam trip diagnostic system (base on bunch-by-bunch system)
Bunch-by-bunch and turn-by-turn
Time domain and frequency domain
Tune in three dimensions
Beam Current：last turn
Beam current in the end：800 turn
Bunch current
Bunch RMS/mm
Position X
FFT X
Position Y
FFT Y ① ③ ⑤ ⑦ ② ④ ⑥ ⑧

10. Some examples Beam trip statistics
Almost all the tripping can be distinguished as the following types:
RF system (including suspected SC cavity trip)
beam instability (failure of the feedback system)
power supply system
Other hardware systems trip resulted in beam lost, such as control system, magnet system, cryogenic system and so on.
Beam trip statistics 10

11. RF trip
voltage
The beam longitudinal oscillation
beam 11

12. Multipacting or stress jump?
RF “f-phase” trip
RF “f-phase” trip occurred frequently when the electron beam >700mA during the past year.
RF forward power
RF reserved power
RF voltage
RF phase
Beam
f-phase jump
Loading angle
Piezo Driver
The loading angle was found jumped sometimes at the end of May 2017, the beam lost when the frequency control loop can not catch up.
Multipacting or stress jump? 12

13. This SC Cavity has just been changed as the f-phase trip is one of the important reasons.13

14. Beam instability Feedback system: Phase shift
Bunch current
Not uniform
Transvers oscillation
Tune X
Feedback system:
Phase shift
the damage of the power amplifier
Bandwidth limited (high current with more bunches)
Undergoing: digital & new feedback kicker

15. Power off with Alarm: easy to find
Power supply
Power off with Alarm: easy to find
Unstable: difficult to find sometimes
It occued a dozen times that the positron and electron were tripped at the same time in the January 2017 and must be caused by the unstable state of the public magnet power supplies (Q1A or ISPB). 15

16. Check the Q1A power output through the inspection system.
Not matching with ISPB
Beam trip by ISPB
Check the Q1A power output through the inspection system.
See nothing: slow system vs. fast change 10ms.

17. Solution: tighten the power supply terminal block!
Knob Q1A (I -15A)
Solution: tighten the power supply terminal block!

18. Summary
Keeping high availability during a long operation period is quite challenging.
The 16-channel oscilloscope and the beam trip diagnostic system are powerful failure analysis tools.
We are training the operators to have capability to use the failure analysis tools.

19. Thanks for your attentions !