Recent Experimental Studies of the Electron Cloud at the Los Alamos PSR

Transcription

1 Recent Experimental Studies of the Electron Cloud at the Los Alamos PSR Robert Macek, 9/11/01 - KEK Workshop Co-authors: A. Browman, D. Fitzgerald, R. McCrady, T. Spickermann and T. S. Wang 1

2 Outline Background: Electron Cloud Issues for PSR Electron Cloud Diagnostics at PSR Biased collection plates Harkay-Rosenberg RFA Electron Sweeping Detector Results from the Electron Sweeper Electron signal Electrons surviving a long gap Recovery from gap clearing Electron Cloud Measurements in a Weak Solenoid Some Remaining Puzzles Summary and Conclusions 2

3 Some Important Electron Cloud Issues at PSR The strong, fast, transverse instability at PSR is almost certainly e-p, based on the many observed characteristics of the unstable proton beam motion See Macek etal, FOAB007 PAC2001 and references therein, also ICANS-XV paper However, the origin and important characteristics of the electron cloud are less well understood Key issue: can we make a compelling case that electron suppression by TiN coatings and/or solenoids will cure the instability? TiN coatings nicely suppress the copious e s striking the wall at the end of each bunch ( prompt e s ), but are these the one that drive the instability? Prompt e s striking the wall are strongly dependent on beam intensity, beam losses and vacuum pressure but this is not reflected in the instability threshold behavior, why? Trailing edge multipactor generates many electrons but what fraction survives passage of the gap to be captured by the following beam pulse? Is multi-turn accumulation of electrons significant? The electron density in the beam (neutralization) is the critical factor for e-p dynamics but is difficult to measure directly 3

4 Mechanism #1: trailing edge multipactor WC41 Electron born at wall from say losses Beam E-Detector x 4 Energy gain is possible in wall-to-wall traversals on trailing part of beam pulse Energy gain in one traversal is high enough for multiplication 4 Bk87, p111

5 Mechanism #2: production by captured electrons Production of Secondary Electrons captured e- Secondary electrons Tertiary electrons. Proton Beam Bunch Vacuum Chamber Wall 5

6 Electron Cloud Diagnostics at PSR Available electron diagnostics in PSR provide valuable information that helps constrain simulations but do not directly measure beam neutralization DC-biased collecting plates are slow and perturb the beam/wall environment but are our only device that can be used inside magnets ANL Retarding Field Analyzers (RFA) measure the electrons striking the wall The pulsed electron sweeping detector comes the closest in that it can measure electrons in the pipe at various times during passage of the gap electron line density at end of gap is a lower bound on electrons captured by the next passage of the beam pulse also can be used to clear the gap of electrons (locally) 6

7 Retarding Field Analyzer Described in R. Rosenberg and K. Harkay, NIM A 453 (2000) p LANL augmentation is fast electronics (~80 MHz) on the collector output Minimal perturbation of beam/wall environment Simplified RFA Installation Sketch Collector Repeller Screen R1 100 K C1 10 microfarads C2 10 microfarads Collector Bias In +45 Volts ED Signal Out RepellerBiasIn +25to-250Volts Use of repeller permits collecting a cumulative energy spectrum Obtain data on e-flux, time structure and energy spectra Beam Pipe ED ~1.9 cm dia. Measures electrons striking the wall, not electrons remaining in the pipe Three Slots in Beam Pipe Total Slot Area~1.05 square cm 7

8 Electron signals from RFA in straight section 4 2 Beam Pulse Electron Signals V rep =+25V 1.5 Amplitude (V) 1 V rep =-1V V rep =-30V Beam Pulse 0.5 V rep = - 275V V rep = - 300V Time (ns) Signals averaged for 32 beam macropulses, ~ 8 µc/pulse beam intensity, device is labeled ED42Y, Transimpedance = 3.5 kω, opening ~1 cm 2 8 Bk95, p6-12

9 Electron energy cumulative spectrum (3D profile) ED02x, ~8 µc 9 Bk 95, p 6-12

10 RFA signals in a single pass experiment 6.8 µc beam pulse in the extraction line Single pass electrons vs beam intensity (log-log plot) Y = 4E-06X Amplitude (V) Beam Electron ROED01Y amp (v) Time (ns) Intensity (µc/pulse) RFA electron signal is very similar to signals in the ring wrt e-flux, time structure, energy spectra and dependence on beam intensity Of the two mechanisms considered, only trailing edge multipactor can produce the signals observed in this experiment 10 Bk xx, p yy

11 Electron-sweeping Detector Layout Collector Repeller Grid Slots & Screen Pulsed Electrode 11

12 Picture of installed electron sweeper ED42Y E-sweeper, ES41Y 12

13 Electron Sweeper Collection Region 0.05 Acceptance of New Detector-α=75 (Particles inside blue lines hit detector region-v=-100 volts) Detector (V=0) Pipe (V=0) Y(meters) Plate (V=-100) Accepted fractional area= X(meters) 13

14 Sample Electron Data from Electron Sweeper Signals have been timed correctly to the beam pulse (wall current monitor) Prompt electrons strike the wall peak at the end of the beam pulse; basically acts a large area RFA until HV pulse applied Beam Pulse HV pulse Note~10nstransittimedelay between HV pulse and swept electron signal is expected Swept electron signal is narrow (~10 ns) with a tail that is not completely understood May be due to secondaries created at ground screen, walls of slots and repeller screen Reduced by higher repeller voltage Electron Signal Swept electron signal Bk 98, p µc/pulse, bunch length = 280 ns, 30 ns injection notch, signals averaged for 32 macropulses, repeller = - 25V, HV pulse = 500V 14 Bk 98, p 51

15 Prompt and Swept Electrons at Two Intensities prompt electrons strike the wall at the end of the beam bunch Signals taken at last pulse in the ring Swept electron signal taken at 50 ns after end of beam pulse Note the strong dependence of prompt signal on intensity (~ I 7 ) while swept electron signal scales ~ with intensity At larger delays swept e signals are nearly the same size Implies that e s surviving the gap scale ~ with intensity or fractional neutralization is ~ constant! Swept e signal amplitude implies ~10 pc/cm line density at end of gap or a lower limit on beam neutralization of ~ 1% Amplitude (V) µc/pulse 7.4 µc/pulse Prompt e's Swept e's Time (ns) Bk 98, p 47 Signals averaged for 32 macropulses, bunch length = 280 ns HV=500V, repeller = -25V, ES impedance 2.5 kω 15 Bk 98, p 47

16 Electron survival after end of beam pulse Early results from electron sweeper for 5µC/pulse looking just after extraction Peak signal or integral have essentially the same shape curve Long exponential tail seen with ~170 ns decay time Still see electrons after 1 µs Implies a high secondary yield (reflectivity) for low energy electrons (2-5 ev) Amp Vpeak Int τ = 170ns δ eff d = exp c τ me c 2E Implies neutralization lower limit of ~1.5% based on swept electron signal at the end of the ~100ns gap T(ns) 16 Bk xx, p yy

17 Comparison of electron survival curve for two intensities µc (7/13/01) Peak (V) µc (6/21/01) T(ns) 17 Bk xx, p yy

18 Recovery after locally clearing gap of electrons Beam Pulse HV pulse E-sweeper signal 18 Bk xx, p yy

19 Correlations in Recovery Proton and Electron Signals vs. Time 1.0 (Electrons swept out at t=0.) PSR Proton Current Scaled Signal Strength (Volts) Nearby Electron Detector Signal Electron Sweeper Signal Time (nsec)-t=0 at half height of leading edge of electron sweeper HV pulse 19 Bk xx, p yy

20 Picture of Solenoid Section with RFA 20

21 RFA Signals in a Weak Solenoid Field B=10 G B=20 G B=4 G B=0 21 Bk xx, p yy

22 Effect of Solenoid on RFA Signal Amplitude 10 1 Amp(V) B(G) 22 Bk xx, p yy

23 Electron burst phenomenon ED42Y ES41Y Local Loss monitor signal 23 Bk 98, p 53

24 Conditioning effect Threshold Intensity Curves 2000 "Conditioning" effect 10 Threshold Intensity ( µc/pulse) /14/00 Historical Data 4/10/00 4/9/00 4/8/ Buncher HV (kv) 24 Bk95, p125-7

25 1000 Conditioning effect in swept e s? µc (7/13/01) Peak (V) µc (8/24/01) µc (8/24/01) T(ns) Data at 100 ns delay show factor of ~1.7 reduction in electrons surviving the gap over a period of 40 days 25 Bk 98, p 47

26 Summary and Conclusions Fast timing information from both the RFA and electron sweeper has been a valuable tool in understanding the electron cloud in PSR The electron sweeping detector works about as excepted and has provided several important results Electrons decay slowly (~170 ns time constant) after the end of beam pulse implying a high total secondary yield (~0.5) for low energy (2-5 ev) electrons Recovery after clearing the gap of electrons takes several turns and shows that multi-turn accumulation makes a sizeable contribution to electrons striking the wall Electrons surviving the gap are not nearly as dependent on beam intensity as the prompt electrons striking the wall at the end of the beam pulse Data on electrons surviving the gap implies a roughly constant fractional neutralization of ~1-1.5% (lower bound), which is more in keeping with observed threshold behavior of the e-p instability The relationship between prompt electrons striking the wall and those surviving the gap is not completely understood The electron burst phenomenon is an unresolved puzzle at this time A weak solenoid (~20 Gauss) attenuates the electrons striking the wall (RFA signal) by a factor of ~20 or more Open issue: Will electron suppression by TiN coatings of all vacuum surfaces and/or solenoids cure the instability? 26

27 Future Plans Exploit electron sweeping detector to study factors that affect electrons surviving the gap and presumably the beam neutralization Can we explain why the instability threshold is not sensitive to losses and vacuum pressure when it clearly affects the electrons striking the wall? i.e., is the fractional neutralization insensitive to beam losses and vacuum pressure? Install an electron sweeper coated with TiN and measure electrons surviving the gap in coated section Measure effect of bellows, ceramic breaks, and TiN coating on both prompt electrons and those surviving the gap Measure the electrons surviving the gap when small amounts of beam are introduced in the gap Try to discover the cause(s) of the electron burst phenomenon Attempt to observe coherent transverse motion of electron cloud during unstable beam motion Use a combination of TiN coatings and solenoid windings to suppress electrons in a significant fraction of the ring in an attempt to raise the instability threshold 27

28 PSR Layout Merging Dipole Stripper Foil H- Beam Matching Section Final Bend Extraction Line Skew Quad Circumference = 90m Beam energy = 798 MeV Revolution frequency =2.8 MHz Bunch length ~ 250 ns (~63 m) Accumulation time ~ 750 ms ~2000 turns C Magnets Bump Magnets H-/H0DumpLine rf buncher 28