BESIII Collaboration Meeting in Winter of 2015 BEAM ENERGY MEASUREMENT SYSTEM

Transcription

1 BESIII Collaboration Meeting in Winter of 2015 BEAM ENERGY MEASUREMENT SYSTEM Nickolai Muchnoi Budker INP, Novosibirsk December 12, 2015 Nickolai Muchnoi BES-III Winter Collaboration Meeting December 12, / 38

2 TALK OUTLINE 1 BEMS history & principles of operation 2 Energy scale calibration 3 BEMS-2015: scale calibration 4 BEMS-2015: beam energy determination 5 Conclusion Nickolai Muchnoi BES-III Winter Collaboration Meeting December 12, / 38

3 BEMS HISTORY Project was started in 2008 First tests and ψ(2s) scan December, 2010 τ mass measurement experiment December, 2011 Continuous operation 2012 Laser met problems 2013 Laser repair, new ZnSe vacuum windows BEMS test with a new laser May, 2015 Nickolai Muchnoi BES-III Winter Collaboration Meeting December 12, / 38

4 Inverse Compton Scattering electron: ε0, γ=ε 0 /m photon: ω electron: ε θ ω θ ε photon: ω 0 Scattering parameters are u and κ: u = ω ε = θ ε θ ω = ω ε 0 ω ; u [0, κ] ; κ = 4ω 0ε 0 m 2. Scattering angles: γθ ω = κ/u 1; γθ ε = u κ/u 1. Maximum energy of scattered photon (θ ω = θ ε = 0): ω max = ε 0κ 1 + κ. ( ) Initial electron energy: ε 0 = ω max m2 m ωmax. 2 ω 0 ω max 2 ω 0 Nickolai Muchnoi BES-III Winter Collaboration Meeting December 12, / 38

5 Accurate energy scale transfer: ev MeV GeV IR optics, 10P20 CO 2 laser line: ω 0 = ev γ-lines from excited nuclei as a good reference for ω max : 137 Cs τ 1/ y E γ = ± kev 60 Co τ 1/ y E γ = ± kev E γ = ± kev 208 Tl τ 1/2 3 m E γ = ± kev 16 O E γ = ± kev High energy physics scale 1 : J/ψ ± ± MeV ψ(2s) ± ± MeV 1 Final analysis of KEDR data, Physics Letters B 749 (2015) Nickolai Muchnoi BES-III Winter Collaboration Meeting December 12, / 38

6 BEMS layout at the North BEPC I.P. positrons electrons R2IAMB HPGe R1IAMB 2.5m 3.25m 3.75m m 0.4m Laser Lenses Size of HPGe detector D 4 cm Distance between HPGe and γe + /γe scattering area L 8 m Beam orbit angle should be zero within θ D/L ±2.5 mrad If θ is outside these limits, measurements are impossible! THIS IS 1 BEMS PROBLEM: NO DATA = NO MEASUREMENT! Nickolai Muchnoi BES-III Winter Collaboration Meeting December 12, / 38

7 BEMS SUBSYSTEMS Laser & optics system provides laser transportation and necessary focusing to the interaction area. Control system provides change of laser direction to electron or positron beam, control over additional moving shield 2, tune (maximize) the rate of backscattered photons. It uses DAQ system counting rates as a feedback signal. DAQ system reads HPGe data from MCA, saves the raw data to disk. Uses Control system status to distinguish electron/positron records. ALL RAW DATA IS AVAILABLE! On-line analysis system provides online beam energy determination results and writes them to the BEPC database. O-line analysis role is to make various checks and get better results. 2 Up to 18 cm of lead shielding was installed to suppress beam background! Nickolai Muchnoi BES-III Winter Collaboration Meeting December 12, / 38

8 1 BEMS history & principles of operation 2 Energy scale calibration 3 BEMS-2015: scale calibration 4 BEMS-2015: beam energy determination 5 Conclusion Nickolai Muchnoi BES-III Winter Collaboration Meeting December 12, / 38

9 Absolute energy measurements by HPGe spectrometers Practical experience has been gained in the eld of nuclear spectroscopy. Idaho group recommendations for precise absolute measurements: use more than one spectrometer simultaneous and unidirectional measurement of calibration lines and energies under investigation perform energy calibration in a narrow range instead of polynomial extrapolation of the whole scale avoid using m 0 c 2 or 2m 0 c 2 values for determination of energy dierence between photo-peak and escape-peaks avoid using pulsers for calibration Nickolai Muchnoi BES-III Winter Collaboration Meeting December 12, / 38

10 Our approach is dierent cause the range where we work is rather wide. So we'll try to: nd an appropriate function to describe the total total absorption peak shape; check that the parameters of this function have a smooth energy dependence; use BNC PB-5 precise amplitude pulser with declared integral linearity as small as 15 ppm. Nickolai Muchnoi BES-III Winter Collaboration Meeting December 12, / 38

11 HPGe energy response function 0 < x < + : exp { x2 } 2σ 2 f(x) = A K 0 K 1 σ < x 0 : C + (1 C) exp { x2 } { 2(K ( 0 σ) 2 x < x K 0 K 1 σ : C + (1 C) exp K 1 K 0 σ + K )} 1 2 A amplitude, x = 0 line energy, σ normal width, K 0 σ width from-the-left modication, K 1 K 0 σ exponential low-energy tail, C is for low-angle scattering of γ-s on their way to detector. Nickolai Muchnoi BES-III Winter Collaboration Meeting December 12, / 38

12 HPGe energy response function Amplitude [a.u.] gauss another gauss exponent tail Compton edge (ω-ω 0 ) in units of σ Nickolai Muchnoi BES-III Winter Collaboration Meeting December 12, / 38

13 6129 kev peak (2011 data) O ( kev) χ 2 /ndf = 61.8/ E γ, kev Nickolai Muchnoi BES-III Winter Collaboration Meeting December 12, / 38

14 HPGe energy resolution (2011 data) σ E = σ εf E ε electron-hole creation energy in Ge, F Fano factor 0.25 / E, % σ E 0.20 reference lines pulser lines other lines E γ, kev Nickolai Muchnoi BES-III Winter Collaboration Meeting December 12, / 38

15 Peak shape widening, K 0 (K 1 = ) K 0 1.6, K, Compton 1 K 0 Compton,% E γ, kev Nickolai Muchnoi BES-III Winter Collaboration Meeting December 12, / 38

16 Wide-range scale calibration (2011 data) E FIT - E REF, kev 0.2 reference lines pulser lines other lines Nickolai Muchnoi BES-III Winter Collaboration Meeting December 12, / 38

17 1 BEMS history & principles of operation 2 Energy scale calibration 3 BEMS-2015: scale calibration 4 BEMS-2015: beam energy determination 5 Conclusion Nickolai Muchnoi BES-III Winter Collaboration Meeting December 12, / 38

18 BEMS test in May, 2015: spectrum example Electrons: [09:08:46 10:44:12] Live time: 0 hours 44 min 20 s E γ, kev Nickolai Muchnoi BES-III Winter Collaboration Meeting December 12, / 38

19 BEMS test in May, 2015: calibration lines t Cs ( kev) χ /ndf = 58.5/ Co ( kev) χ /ndf = 65.8/68 60 Co ( kev) χ 2 /ndf = 86.2/ E γ, kev E γ, kev E γ, kev Nickolai Muchnoi BES-III Winter Collaboration Meeting December 12, / 38

20 BEMS test in May, 2015: energy scale E FIT 0.4 reference lines pulser lines 0.2 other lines E REF, kev E γ, kev Nickolai Muchnoi BES-III Winter Collaboration Meeting December 12, / 38

21 BEMS test in May, 2015: resolution and K 0 / E, % σ E reference lines pulser lines K 0, K 1, Compton K other lines 1.4 Compton,% E γ, kev E γ, kev Nickolai Muchnoi BES-III Winter Collaboration Meeting December 12, / 38

22 BEPC orbit inuence example: GOOD (e ) & BAD (e + ) Electrons: [08:50:05 09:02:07] Live time: 0 hours 7 min 43 s Positrons: [09:32:09 09:43:28] Live time: 0 hours 4 min 32 s E γ, kev E γ, kev Nickolai Muchnoi BES-III Winter Collaboration Meeting December 12, / 38

23 Energy equivalent of the MCA Channel 5445 CH linear energy equivalent, kev electrons positrons Apr 29 Apr 30 May 01 May 02 May 03 May 04 May 05 Nickolai Muchnoi BES-III Winter Collaboration Meeting December 12, / 38

24 Scale correction (by pulser) for MCA Channel 5445 CH pulser energy correction, kev electrons positrons Apr 29 Apr 30 May 01 May 02 May 03 May 04 May 05 Nickolai Muchnoi BES-III Winter Collaboration Meeting December 12, / 38

25 Energy resolution at 2000 kev CH energy resolution, % electrons positrons Apr 29 Apr 30 May 01 May 02 May 03 May 04 May 05 Nickolai Muchnoi BES-III Winter Collaboration Meeting December 12, / 38

26 Peak shape parameter K 0 at 2000 kev extrapolation to CH K electrons positrons Apr 29 Apr 30 May 01 May 02 May 03 May 04 May 05 Nickolai Muchnoi BES-III Winter Collaboration Meeting December 12, / 38

27 Conclusions about energy scale calibration Absolute energy scale is based on γ-quanta emitted by excited nuclei. During BEMS-2015 test 137 Cs and 60 Co γ-sources were used. With only three lines the statistical uncertainties for all calibration coecients are rather large. The arbitrary uncertainty of scale calibration is at the level of 10 4, it can be signicantly reduced by adding 2614 kev line ( 208 Tl), as it was done before, during routine BEMS operation in Unstable beam orbit leads to large background uctuations. This is the main reason of instabilities in the calibration procedure. Special care should be paid to satisfy BEMS requirements for beams orbits in the laser interaction area. Nickolai Muchnoi BES-III Winter Collaboration Meeting December 12, / 38

28 1 BEMS history & principles of operation 2 Energy scale calibration 3 BEMS-2015: scale calibration 4 BEMS-2015: beam energy determination 5 Conclusion Nickolai Muchnoi BES-III Winter Collaboration Meeting December 12, / 38

29 Compton edge tting Add additional Gaussian spread σ s to the peak-shape, which comes from electron beam energy spread (x = ω ω max, σ m K 0 σ, K 1 = ): { S(x) = 1 2 2π 1 exp σ 2 + σs 2 + C ( x ) σ erfc + (1 C)K ( 0 exp 2σs σ 2 m + σs 2 ( x 2 2(σ 2 + σ 2 s) x 2 2(σ 2 m + σ 2 s) ) ( erfc ) ( erfc The convolution to describe the edge of Compton spectrum: xσ σ s 2(σ 2 + σ 2 s) ) + xσ m σ s 2(σ 2 m + σ 2 s) ) } E(ω, ω max ) = + S(ω ω max )dω + Background ω Nickolai Muchnoi BES-III Winter Collaboration Meeting December 12, / 38

30 Edge Fit 600 K = 1: χ 2 0 /NDF = 314.2/ ω max = ± 0.15 kev 400 what happens if K = Electrons: [00:05:25-00:17:27] Live-time: 0 hours 7 min 32 s E γ, kev K 0 = 1.48 ± 0.15: χ /NDF = 304.3/295 ω max = ± 0.15 ± 0.22 kev E γ, kev Nickolai Muchnoi BES-III Winter Collaboration Meeting December 12, / 38

31 Copy of tting output Simple Edge Fit: Range from to kev E_beam = MeV W_max = kev Edge amplitude : ± Edge slope: ± Edge wmax, kev: ± Backgrond level: ± Edge width, kev: ± Background slope: ± χ2/ndf = 314.2/307 Probability: Complex Edge Fit: Range from to kev Amplitude = W_max = kev HPGe resolution = kev HPGe K0 = Spread = kev Edge wmax: ± 0.15 ± 0.22 kev Beam σe impact: 2.73 ± 0.28 ± 0.18 kev Edge amplitude : ± Backgrond level: ± HPGe resol, kev: ± Background slope: ± HPGe K0 : ± Compton slope: ± χ2/ndf = 304.3/295 Probability: Wmax: ± 0.15 kev (symmetric fit) Wmax: ± 0.26 kev (asymmetric fit) Wmax: ± 0.32 kev (linear scale error) Wmax: ± 0.32 kev (spline correction ) electron Beam Energy Determination: BEPC beam energy = ± MeV was taken from database Measurement time from :05:25 to :17:27. BEMS beam energy = ± MeV (SR correction to IP MeV was added) BEMS beam spread = 682 ± 83 kev Nickolai Muchnoi BES-III Winter Collaboration Meeting December 12, / 38

32 BEMS results: electron beam energy. E 1.5 MeV Beam energy, MeV BEPC energy: electron beam BEMS energy: electron beam χ / ndf / 237 p ± May 01 May 01 18:00 May 02 May 02 18:00 May 03 May 03 18:00 May 04 May 04 18:00 May 05 Nickolai Muchnoi BES-III Winter Collaboration Meeting December 12, / 38

33 BEMS results: positron beam energy. E 0.9 MeV Beam energy, MeV BEPC energy: positron beam χ 2 / ndf / 212 p ± BEMS energy: positron beam May 01 May 01 18:00 May 02 May 02 18:00 May 03 May 03 18:00 May 04 May 04 18:00 May 05 Nickolai Muchnoi BES-III Winter Collaboration Meeting December 12, / 38

34 1 BEMS history & principles of operation 2 Energy scale calibration 3 BEMS-2015: scale calibration 4 BEMS-2015: beam energy determination 5 Conclusion Nickolai Muchnoi BES-III Winter Collaboration Meeting December 12, / 38

35 Conclusion BEMS is ready for new experiments. Whenever accurate beam energy determination is important in a particular experiment, the following steps should be performed: Contact BEMS group for joint planning of experiment. Contact BEPC team to provide special care on orbits at the North IP (even with possible luminosity loss). Perform fast scans of J/ψ or(and) ψ(2s) resonances for better understanding of systematic uncertainties. Do oine study of BEMS data to check everything prior and along with BES-III data acquisition runs. Long-term BEMS operation without performing the above steps leads to rise of low quality data records. Analysis and understanding of such data becomes more dicult, resulting in increase of systematic errors. Our experience convinces that BEMS should not operate in cases when beam energy is not a matter of interest. THANK YOU! Nickolai Muchnoi BES-III Winter Collaboration Meeting December 12, / 38

36 Orbit radius oscillations (BPR) from BPM signal Nickolai Muchnoi BES-III Winter Collaboration Meeting December 12, / 38

37 Orbit radius oscillations (BER) from BPM signal Nickolai Muchnoi BES-III Winter Collaboration Meeting December 12, / 38

38 Orbit radius oscillations (BER) from BPM signal Most probable explanation for the observed oscillations is the oscillations in BEPC guide eld, where frequencies are the multiples of AC line frequency. If so, this denitely leads to average energy oscillations. Long-time average distribution of the electrons energies is no more a normal distribution. If so, the edge tting procedure becomes incorrect, leading to systematic shift of results. Nickolai Muchnoi BES-III Winter Collaboration Meeting December 12, / 38