PEP-II Status. U. Wienands, PEP-II Run Coordinator for the PEP-II team

Transcription

1 PEP-II Status U. Wienands, PEP-II Run Coordinator for the PEP-II team

2 Outline of Talk Run 4 Synopsis Machine tuning & improvements Issues encountered during Run 4 Other improvements and MD items Outlook: the remainder of Run 4 2

3 Run 4 Synopsis PEP is now delivering twice the rate of Run 3 >4 times CDR rate Over 60/fb delivered in Run 4, over 200/fb tot Key improvements: trickle-charge (continuous injection) both rings Raise peak Lumi to 8.34E33 (currents, ß*, orbit) Aggressively track down & reduce rf trips. MD program focused on near-term improvements + strategic studies needed for upgrades. 3

4 Run 4 Delivered Luminosity 4

5 5

6 Running 7-day Integral (2004) 6

7 Run Time Accounting, Run 4 BaBar up from 56% to 60% Tuning/Fill down from 22% to 20% Downtime down from 16 19% to 14% 7

8 9

9 Machine Tuning and Improvements LER & HER trickle charge Steering of the HER Lowering ß y * of the HER (12 >11 mm) Raising of the HER rf voltage Use HER Rf 4-1 Raising the number of bunches & beam currents 10

10 LER Trickle Charge PEP issues mostly resolved last fall BIC, diagnostics software Extended BaBar trickle test late Oct. 03 followed by analysis of data quality Trickle in production beginning Dec. 2,03 Daily luminosity up from 250 to 350/pb 15% is direct trickle gain, remainder is increased efficiency (less trips) Dec. a particularly stable month for rf. 11

11 LER only Trickle Day 12

12 LER Trickle Background Monitor FFT of Injection- triggers (ν s harmonics if off energy) Injection-generated triggers ( also avail. as EPICS PV) 13

13 HER Trickle Charge 1st test in Feb. 04 significant reduction in injection background not at full beam current, high radiation spikes 2nd test mid-march 04 Radiation spikes no longer an issue due to cleanup work in injector Break-even luminosity reached. BaBar state machine now ready for double trickle Double-trickling operationally since March

14 Double-Trickle Gain LER-only trickle av-to-pk = 87% Lum (10^30) 6000 Lumi stable beam 7-Dec-03, 16:00 24:00 frac_on:0.94 av_pk: av_lum: av_to_pk:0.87 tau: Double trickle av-to-pk = 99% => 12% gain 0 Lum (10^30) Time (m) Lumi stable beam 6-Apr-04, 00:00 08: av_lum: av_to_pk:0.99 av_pk: frac_on: Time (m) 15

15 Almost Best Day of PEP-II The best day actually had 669/pb delivered. 16

16 Fill Patterns & Beam Currents At the beam-beam limit need to raise # bunches and beam current to gain lumi. Need to run by-2 (4.2 ns) bunch spacing: How will electron-cloud effects behave? Will luminosity scale with # bunches? What is the effect of parasitic crossings? How will HOM heating behave? (only the last expected to be better) 17

17 Spec. Luminosity vs # Bunches Fixed bunch current in both rings Fixed by-2 pattern with mini trains Vary train length, thus # of bunches 18

18 Spec. Luminosity at 6.3 and 4.2 ns Substantial initial luminosity drop when switching to by-2 pattern. Mostly recovered after a few weeks by 3 (6.3 ns) by 2 (4.2 ns) 19

19 Raising # Bunches & Beam Currents Bunch currents: 1.5 ma (LER) 0.9 ma (HER) fairly constant 20

20 PEP Fill Patterns 1415 bunches trains of 15, 4.2 ns 1550 bunches trains of 33, 4.2 ns Note: 1st and last bunch have lower luminosity: only one parasitic xing Pilot bunches (no parasitics) have even less luminosity! 21

21 Luminosity vs I HER *I LER 22

22 HER Steering Steer the HER flat except for IR 2 this has worked historically IR 2 steering in the past has caused difficulties with coupling Motivation lower vertical dispersion better acceptance better agreement of the lattice with the design 23

23 Orbit (absolute) Y dispersion ß x /ß x,design Coupling (C 12,n ) 24

24 HER ß y * Reduction We used 60% of a knob designed to bring ß* down to 10 mm Before ß y * reduction ß y * 12.3 mm ß x * 28 cm Benefit for luminosity HER beam stronger 3% increase in spec. lumi. Before STDZ After ß y * reduction ß y * 11.0 mm ß x * 27 cm After STDZ ß y * 10 mm ß x * 25 cm ß x * lower due to larger global ß x beat. Significant tuning work required post-stdz to regain luminosity benefit 25

25 Coupling after ß y * Reduction 26

26 Lowering HER ß y * (cont d) To be able to meaningfully reduce ß y *, we are raising the HER rf voltage. At 18 MV we expect 11 mm bunch length. The rate of rise is limited by increased rf trip rate. 27

27 Issues encountered IR 2 vacuum and background issues Rf issues LER orbit & beam optics issues Bunch-by-bunch feedback issues 28

28 The IR 2 Vacuum Problem Late March, we regn d the IR 2 NEGs Pressure and background had been on rise At Run 4 startup, NEGs regen d before 1 st beam After this, LER backgrounds and pressures much worse than before Also, long time constants ( hours) observed Strongly bunch-length dependent Scrubbing very slow, almost 2 weeks NEGs being heated by HOMs to several 100 F 29

29 LER Vac. Gauge close to IR 30

30 LER Vacuum -50 m 31

31 NEG Temperature at -28 m In essence, the NEG is being constantly regen d with beam! 32

32 NEG Pump Chamber TC # hottest Collimator 33

33 Rf Issues Rf has been holding up better than in Run 3 thanks to significant effort of the RFTF The primary issues/causes of trips endemic failure (again) of SCRs in the HVPS phase drifts (esp. LER 4-5) & other board issues various tuner issues klystron issues misc. IOC issues 34

34 PEP Rf Beam Aborts Run 3 Run 4 HER rf trips: 5.0/day LER rf trips: 2.0/day Total: 7.0/day HER rf trips: 2.1/day LER rf trips: 0.9/day Total: 2.3/day incl. l. inst. 3.7/day Despite the overall improvement, # rf trips actually increased significantly after Dec

35 LER Orbit Issues Steer the ring without the IR, as has worked before. LER IR steering perilous due to BPM offsets This caused a large drop in spec. luminosity Root cause was a large beat in ß x, also, dispersion still >50 mm rms. No obvious quick fix => backed out Presently using MIA to try finding a correction strategy. 36

36 LER Steered Orbit 37

37 LER ß x after Steering 38

38 Other improvements and MD BPM improvements Modeling improvements (incl. ORM, C 12 ) Transverse grow-damp measurements Longitudinal grow-damp measurements New RF 8-1 klystron New rf IOCs using PowerPC controllers Injection Crossing-angle experiment 39

39 BPM Improvements BBA on all possible IR 2 BPM/quad pairs Offsets from this BBA now in DB Auto-calibration every 8 hours after a beamdump Timing database updated New procedure to measure timing during delivery See talks by Smith & Ross 40

40 Online Model Improvements Updated online models for both rings Survey data for IR 2 elements More accurate parametrization of QD1 Can run a model against machine config. & orbit Model runs on pepoptics => can use MAD & new version of DIMAD. C 12 coupling analysis now online 41

41 ß_y (m) HER Model vs Measurement(y) Distance (m) ß_y (m) Distance (m) ß_y (m) ß/ß (%) 50 0 green: data, blue: exc. data, red: fitted model HER, ß_y, S240, after steer, new design model, 8-Jan Distance (m) 42 Z_waisty : beta_waisty:

42 History of ß Functions at the IP 43

43 Orbit-Response Measurements ORM (LOCO) running on pepoptics HER and LER data taken ß function analysis consistent with phaseadvance analysis More work needed to understand quadstrength factors => presentation by C. Steier 44

44 HER ORM ß/ß design ? x * = 26.6 cm ? y * = 1.02 cm

45 HER ORM Magnet strengths Relative to design strengths. E(HER) = GeV

46 Multibunch Stability Grow-damp measurements HER: longitudinal & transverse HER transverse: good margin in x, strong nonlinearity in y, longitudinal: only 20 30% margin for lowlying long. modes Low-group-delay Woofer significantly better. As a result of the HER measurements, we raised priority on the new Woofer 47

47 HER TFB Damping (x) HER X damping Data suggest stability up to 2 3 A beam current by-2 fill pattern. 48

48 Crossing Angle Experiment Xing angle or not for the IR upgrade in 2005? Measurements done: Sensitivity of x to beam-beam Sensitivity of Lumi to xing angle vs currents Results: L vs x separation becomes narrow at high I L vs x becomes somewhat narrower at high I 49

49 Xing Angle Expt: Lum. vs X Simulation by I. Narsky using 3-d code by Y. Cai 50

50 Injection Orbit Feedbacks With both rings trickling injection problems are highly visible. We are implementing feedback systems to stabilize the beams LER injecting beam now working, HER to follow requires BPM hardware changes Ring-orbit stabilization under development tricky because of modeling issues (Note: we do not let the GOFs steer at injection) 51

51 LER Injection Orbit Feedback 52

52 Rf Improvements New PowerPC IOC processors Old 68k units above 80% CPU load, causing stalls New LLRF boards New SLAC klystron for HER

53 New SLAC Klystron in

54 Outlook: The remainder of Run 4 Lattice improvements Further ß y * reduction in both rings Complete LER steering (MIA work) Fix HER ß x beat Beam-current increases 55

55 ß* Limits Bunch lengths > 1 cm LER: 3.8 MV (achieved) HER: 18 MV (tough!) ß y * 1 cm feasible & meaningful tested knobs exist for both rings Lowering ß x *?? not meaningful if beam-beam limited in both planes Maintain matched beam sizes 56

56 Global Ring Issues Knob to reduce global HER ß x beat exists good in 1 st order but beam life time reduced need to investigate higher-order effects Correction of global ß x beat in steered LER 1 st MIA solution has some very strong quad changes => may be problematic in practice source of ß beat not yet clearly identified 57

57 HER Beam current limits Rf power limit: 1600 ma Stability: 1400 at present, 1600 ma with new woofer LER Tested up to 2400 ma, likely higher limit with llrf tuning rf power limit 3000 ma thermal issues with TFB, LFB kickers 58

58 July 2004 Parameters (Goal) Parameter HER LER I (ma) n bunches ßy* (cm) 1 1 ξ x ξ y V rf (MV) 18 4 Lum (cm -2 s -1 )

59 Acknowledgments The work reported on in this talk has been done by many, in particular: F.-J. Decker, A. Fisher, R. Iverson, W. Kozanecki, S. Novokhatski, M. Ross, J. Seeman, M. Stanek, M. Sullivan, P. Tenenbaum, D. Teytelman, A. Wolski, C. Steier, M. Woodley, G. Wormser, G. Yocky, et al. 60