Electron Beam Properties and Instrumentation MOLLER Director s Review, Jan. 14, 2010 Mark Pitt, Virginia Tech This talk will focus on the electron beam properties and beam instrumentation requirements that are needed to achieve the MOLLER experiment systematic error goals. These aspects of the charge are touched on by this talk: Working group members and others who contributed to this topic: J. Benesch, G. Cates, Y. Kolomensky, K. Kumar, D. Mack, J. Musson, K. Paschke, M. Pitt
Parity Violation Experiments at Jefferson Lab The CEBAF accelerator has been delivering high quality polarized electron beams for parityviolating experiments since 1998: CW beam Oct 13 QE dropped by factor of 2 Nov 9 Currents ~ 20 80 A at up to 86% stable beam polarization Very stable beam properties (small position, angle, intensity, and energy fluctuations) Steady improvements in polarized source (lasers and gun/gaas crystal technology) and close collaboration between experimenters and accelerator staff on minimizing helicity-correlated beam properties Completed experiments: HAPPEx I/II/III, G 0 forward and backward angle, 6 GeV PVDIS photocathode anode Laser -100 kv e-
MOLLER Systematic Error Budget source of error % error absolute value of Q 2 0.5 beam second order 0.4 longitudinal beam polarization 0.4 inelastic e-p scattering 04 0.4 elastic e-p scattering 0.3 beam first order 0.3 pions and muons 0.3 transverse polarization 0.2 photons and neutrons 0.1 Total 10 1.0 What capabilities are needed to meet the electron beam related systematic error goals? Goals for: Electron beam random fluctuations (ie. jitter in beam angle, position, energy, intensity) Electron beam property measurement resolution (ie. How well do we have to make RELATIVE measurements of beam properties?) Size of helicity-correlated electron beam properties (controlling these involves both accelerator and polarized source next talk by Gordon Cates)
Helicity-Correlated Beam Properties: False Asymmetry Corrections Why is measurement and control of relative beam properties critical for parity violation experiments? control of false asymmetries P = P N + P - Y = Detector yield A meas Example: A phys A Y 1 Y i 1 2 P i P i P = beam parameter energy, position, angle, intensity it Y ~ 4.2 ppb / nm x 0.5 nm 1 2Y x false 1 2Y Y x ~ 21 2.1ppb (~ 6% of PVasymmetry) x Need to keep run-averaged helicity correlations of all beam properties p small (< 10 ppb charge, < 0.5 nm position) measure all beam properties accurately enough to keep the error on the corrections small P P P Helicity-correlated beam properties: keep small with careful source/accelerator setup and active feedback when necessary 1 2Y Y P Detector sensitivity: can be reduced for position and angle with symmetrical detector setup
Electron Beam Monitoring Instrumentation in use at JLAB Microwave cavity monitors: Electromagnetic cavity resonant at accelerator RF (1497 MHz) TM 010 measure beam intensity TM measure beam position TM 110 Stripline beam position monitors standard JLAB beam position monitor 4 quarter-wave antennae uses switched electrode electronics (SEE) The availability of both cavity and stripline beam position monitors in Hall A makes redundant beam position and angle measurements possible.
Goals for 1 st order electron beam corrections Beam Assumed Precision of Required Property Sensitivity correction run averaged 1 Y P Y helicity correlation Y P or Y P P P P Systematic contribution Charge 1 ppb / ppb ~1% <10 ppb ~0.1 ppb Asymmetry energy 1.4 ppb/ppb ~10% <0.3 ppb ~0.05 ppb position (on target) 0.85 ppb/ nm ~10% <0.5 nm ~0.05 ppb angle 8.5 ppb/nrad ~10% <0.05 nrad ~0.05 ppb 1 st order corrections: those involving the first moments of the beam properties table assumes a conservative factor of 10 suppression in sensitivity for position and angle after averaging over azimuthal segments Goals are generated by these considerations: Required run-averaged helicity-correlation: it l ti keep correction from each parameter < statistical error (0.6 ppb) Precision of correction: know each correction to ~ 10% of itself
Specifications on Position and Charge Monitor Resolutions 1. Beam position monitors: After regressing out the measured beam properties, the yield will still have remaining random fluctuations from the finite precision of the monitoring instrumentation; this needs to be small relative to the counting statistics width of ~ 200 ppm for 1 khz pairs for a single azimuthal element of the detector. Set goal of 20 ppm; what position monitor resolution is need to achieve this? 1 dy P random 2Y dp 1 position : 8.5 ppb/nm 2 angle : energy : 1 2 1 2 85 ppb/nr pos angle 20 ppm 20 ppm monitor 5 m (471 nr)(10 m) -5 1.4 ppb/ppb 20 ppm (2.9 10 )(4 m) 116 m energy monitor monitor Goal for position monitor resolution: ~ 3 µm for 1 khz pairs 1 2 3 m Note: 16 µm resolution gives <1% statistical broadening for full detector averaging 2. Charge monitors: Yields will be normalized to the charge monitors, but there will be remaining fluctuations due to the finite precision of the charge monitors; Set goal to keep this at 10 ppm or below less than 1% increase of counting statistics width Goal for charge monitor resolution: ~ 10 ppm for 1 khz pairs
Specifications on Beam Position and Charge Fluctuations 1. Beam position/angle/energy fluctuations: Goal is for corrections from these helicity-correlated effects to be less than the statistical ttiti error. This requires Run averaged position difference < 0.5 nm Run averaged angle difference <.05 nr Run averaged fractional energy difference < 0.3 ppb To demonstrate t that t we have achieved these, the random beam fluctuations ti must be small enough to achieve these numbers at 1 after 2 x 10 10 pairs (5000 hours) pos (0.5 nm) 210 10 ~ 70 m angle (0.05 nr) 210 10 ~ 7 R energy (0.3 ppb) 210 10 ~ 40 ppm Goal for position/angle/energy fluctuations: < 70 µm, 7 R, 40 ppm for 1 khz pairs 2. Charge fluctuations: Corrections will need to be made for helicity-correlated charge coupled with nonlinearities in the detector response; to keep this correction below the statistical error requires: Run averaged charge asymmetry < 10 ppb To demonstrate that we have achieved this, the random beam fluctuations must be small enough to achieve these numbers at 1 after 2 x 10 10 pairs (5000 hours) ch arge (10 ppb) 2 10 10 ~1400 ppm Goal for charge fluctuations: < 1400 ppm for 1 khz pairs
Summary of Beam Monitor Resolution and Beam Fluctuation Goals Beam Property Fluctuations for 1 khz pairs* Beam monitor type Monitor resolution for 1 khz pairs Charge < 1400 ppm Charge < 10 ppm energy e < 40 ppm position < 3 µm position (on target) < 70 µm angle < 7 µr *Beam fluctuations at other frequencies can be higher than this. The most problematic other frequency is 60 Hz; the experiment will take data in a scheme synchronized to 60 Hz in order to suppress this noise through averaging.
Beam Position and Charge Monitor Resolutions at 15 Hz Our experience with beam charge/position monitor resolutions is primarily at 15 Hz pair rate. Charge: Recent determination with 18 bit ADCs at 80 A gives ~ 13 ppm for single BCM at 15 Hz ~ 18.8 ppm A BCM2 (ppm) A BCM1 (ppm) Residual (ppm) Stripline BPM monitor resolution measurement Postion: ~ 2.4 m Typical stripline monitor resolutions at 15 Hz ~ 2-3 m Typical cavity monitor resolutions still being determined (but they are less than ~ 1.5 m at 15 Hz from preliminary analyses)
Beam Position and Charge Fluctuations at 15 Hz Typical values at 15 Hz and high (> 40 A) currents are: Position ~ 5 20 m Charge ~ 300 800 ppm Fractional energy ~ 5 10 ppm Typical data during G 0 running in Hall C: charge ~ 436 ppm x ~ y ~ E/E ~ 18 m 6 m 7 ppm
Charge Fluctuations and Monitor Resolution at 1 khz Our experience with jitter is primarily at 15 Hz, but we have some limited experience with higher frequencies from June 2008 Qweak test run; data taken at 15, 125, and 500 Hz. Frequency Dependence of Beam Charge Fluctuations ~ 930 ppm at 15 Hz ~ 904 ppm at 125 Hz ~ 593 ppm at 500 Hz the goal of ~ 1400 ppm charge fluctuations at 1 khz looks achievable Frequency Dependence of Beam Charge Monitor Resolution Only have data at 15 Hz ~ 13 ppm for single monitor Worst case extrapolation to 1000 Hz Assume that t the receiver noise spectrum is white (ie. V/sqrt(Hz) is constant) t) Then the worst case resolution is ~ (13 ppm) 1000 Hz 15 Hz ~ 100 ppm need improvement by ~ factor of 10 to achieve resolution goal of ~ 10 ppm at 1 khz
Beam Position Fluctuations and Monitor Resolution at 1 khz Data from June 2008 Qweak test run; data taken at 15, 125, and 500 Hz. Frequency Dependence of Beam Position Fluctuations for stripline beam position monitor x ~ 6.9 um at 15 Hz x ~ 14.1 um at 125 Hz x ~ 19.2 um at 500 Hz y ~ 10.6 um at 15 Hz y ~ 19.1 um at 125 Hz y ~ 29.3 um at 500 Hz Widths have contributions from both monitor noise and beam position jitter; jitter likely dominates here Take worst cases: Beam position fluctuations: assume all of the width comes from beam position jitter 30 m at 500 Hz makes our goal of 70 m at 1 khz look reasonable Beam monitor resolution: assume cavity monitors achieve goal of 1 m @ 15 Hz (1.5 m shown so far) scale by worst case white noise assumption: sqrt(1000/15) ~ 8 m at 1 khz; goal is ~ 3 m If needed, one can construct cavities of copper-plated stainless steel to replace the existing stainless steel cavities (Q increase by ~ 3)
Summary of Beam Monitor Resolution and Beam Fluctuation Goals Items in red are extrapolations based on measured values; items in green are measured values Beam monitor type charge Monitor resolution for 1 khz pairs < 10 ppm ~ 100 ppm at 1 khz position < 3 µm ~ 8 m at 1 khz Beam Property Fluctuations for 1 khz pairs* charge energy position (on target) angle < 1400 ppm ~ 600 ppm at 500 Hz < 40 ppm ~ 10 ppm at 15 Hz < 70 µm ~ 30 m at 500 Hz < 7 R ~ 1 R at 15 Hz More data on this will be taken during PREX (March 2010, up to 125 Hz) and Qweak (May 2010 onwards up to 500 Hz routinely and occasional tests t up to 1 khz) Biggest improvement needed in charge monitor resolution - see R&D slide for further details Some prospects average multiple l monitors (4 exist in Hall A line now) increase the Q of the charge cavities from their current de-qed state (factor of 3) improved receiver electronics (needs modification to new low noise receiver being tested) change cavities from stainless steel to copper-plated SS (improve Q by factor of ~3)
Slow Helicity Reversal Slow helicity reversals are an important component of a comprehensive strategy to control helicity-correlated correlated beam properties Why use slow reversal: Comparison to two data sets rules out gross problems, at the level lof ~4σ of final error bars Addition of two data sets implies cancellation of subtle problems (at least those susceptible to cancellation under the reversal) Why use more than one: Effectiveness relies on flipping helicity without changing systematic effect... you need the right flip for the specific possible systematic effect E158 Slow Helicity Reversal Results Techniques: Insertable half wave plate most effective for canceling electrical helicity pickup most beam asymmetries are not canceled by IWHP only technique used at JLAB to date g 2 spin precession slight accelerator energy change Double Wien/solenoid spin manipulation in polarized injector
g-2 Spin Flip At 11 GeV, the total precession of the electron s spin relative to momentum ~ 120 Spin flip () can be achieved with modest energy shift of ~ 93 MeV (< 1% change) Energy change small enough that backgrounds, spectrometer optics, etc. should remain very similar SLAC E158 used an In particular, should be effective way of averaging out any helicity- it energy change to correlated beam spot size effects create a g 2 spin flip into End Station A If this were done ~ once per month, we would get ~ 12 g-2 flips over the course of the production run
Injector Spin Manipulation g-2 spin flip involves some accelerator retune a less invasive slow spin flip method is in development for PREx and Qweak 2 Wien filters plus solenoid lens Spin flip is accomplished by +/- 90 o rotation with a solenoid lens focussing in solenoid goes as B 2 so this flip should preserve injector optics with minimal changes Effective way to average out most helicity-correlated beam effects, including spot size If this were done ~ once per week, could get ~ 50 spin flips over the course of the run Two Wien rotations, optimized once then held constant, with +/ 90 degree solenoid rotation
R&D Priorities The major R&D effort in this category will be work on achieving the goal of 10 ppm charge resolution at 1 khz pair rate we need ~ 10 improvement compared to the current worst case extrapolation ti estimate t of ~ 100 ppm There are several short term and longer term strategies for addressing this Short term (can be done parasitically with PREx (March May 2010) and Qweak (May 2010-2012) Increase the Q of the charge cavities from their current de-qued state (500 1500) Take data with the alternate cavity receiver electronics that exists in Hall A and C (Musson design) - perhaps they are better Try the new low noise receivers in development by the RF group (John Musson) Longer term (if further improvement is needed after the above steps) Switch from stainless steel to copper-plated SS cavities (Q increases by factor of ~3) Investigate whether ferrite-core toroid monitors would have better performance Note - R&D project that needs to be started as soon as possible: There is common mode noise from the fact that all the charge monitors use the same local oscillator (LO) as a mixing reference Initial discussions with John Musson indicate that control of the phase and amplitude noise in the LO at this level (10 ppm) is achievable, but it will require a monitoring system to verify it; such a system would be a priority for early development if a microwave cavity solution is adopted
Conclusions Achieving the MOLLER systematic error goals for first and second order helicity-correlated beam corrections requires specifications on these aspects of beam properties and instrumentation: Electron beam random fluctuations at 1 khz pair rate - specifications achievable given conservative extrapolations from existing data - direct measurements of the beam properties at 1 khz pair rate to be accomplished early in the Qweak program (mid-2010) Electron beam property measurement resolution at 1 khz pair rate - position: specification is a factor of 3 below conservative extrapolation from existing data; if needed, improvements could be obtained with copper-plated SS cavities - charge: specification is a factor of 10 below conservative extrapolation from existing data; a staged R&D plan has been developed to address this starting in mid-2010 with the PREx and Qweak runs Size of helicity-correlated correlated electron beam properties - the polarized injector setup aspects of this will be covered by Gordon Cates - second order effects like helicity-correlated beam spot size will be suppressed via two slow spin flip techniques: g-2 spin flip and injector spin manipulation