Beam monitors in J-PARC

Transcription

1 Beam monitors in J-PARC KEK, Accelerator division H. Kuboki Introduction: Beam Monitors in J-PARC BPM system in J-PARC Main Ring BPM gain calibration (Beam Based Gain Calibration (BBGC)) T. Toyama, S. Hatakeyama A, J. Takano, M. Tejima KEK, A JAEA

2 J-PARC Kamioka Tokai, Ibaraki operation (goal) ν detector LINAC 16mA (50mA) 3 GeV RCS 215 kw (1MW) 30 GeV MR 190kW (750W) MLF Hadron hall

3 J-PARC Main ring (MR) Hadron Rapid Cycling Synchrotron (RCS) LINAC MLF Ion Source Neutrino

4 Monitors in J-PARC Beam duct x,y (horizontal, vertical) Profile Position Beam z (beam direction) Bunch Shape Loss Current Time varying information (Turn by Turn etc.) etc.

5 Monitors in J-PARC Profile x,y Beam Position z RCS & inj. BT 54 BPM(COD, turn-by-turn) 3 BPM (RF) 4 BPM (fast) 1 BPM (tune) 2 DCCT / SCT 7 MCT, FCT, WCM 2 IPM 7 MWPM 90 BLM (proportional) 24 BLM (ionization) 20 BLM (scintillator) 2 Exciter 3Ν BT (up to 3N dump) 1 FCT 3 BPM 32 BLM 2 Halo monitor ν BT 1 FCT Abort dump line 2 BPM 1 SEEM 4 BLM T. Toyama Bunch Shape Loss MEBT 8 BPM 6 SCT 5 FCT 4 WS/BSM 4 BLM RFQ DTL/SDTL 29 BPM 18 SCT 47 FCT 4 WS/BSM 53 BLM BSM: beam size mon. A0BT 181MeV L3BT & dumps 17 BPM 3 SCT 4 FCT 4 WS/BSM 30 BLM ACS: under construction Additional devices are in preparation. H0 dump line 1 FCT 48 BPM 11 SCT 5 FCT 24 WS/BSM 38 BLM ACS 400MeV 42 BPM 21 SCT 41 FCT 4 WS 3 BSM 21 BLM 3-50BT 5 FCT 14 BPM 5 SEEM 50 BLM (proportional) 4 BLM (ionization) Bunch Shape Monitor (INR) MR 2 DCCT 7 FCT 2 WCM 186 BPM (COD, turn-byturn) 2 BPM (stripline) 238 BLM (proportional) 36 BLM (ionization) 2 BLM (scintillator) 1 SEEM 5 Luminescence screen 3 IPM 2 Flying wire 2 Exciter Hadron BT 1 SEEM *Monitors not counted for beam transport lines to the utilities, 3N BT, Hadron BT, ν BT

6 Bunch Shape Monitors (LINAC) A. Miura et al. Developed by A.V.Feschenko, P.N. Ostroumov et al, INR, Moscow A. Miura et al., INR 月 J-PARC LINAC に設置中 測定例 (A. V. Feschenko et al.,pac07) 70 ps

7 Profile Monitors Residual gas Ionization Profile Monitor: IPM HV feedthrough K. Satou et al. from T. Toyama Flying Wire Profile Monitor S. Igarashi et al. Multi Ribbon Beam Profile Monitor Y. Hashimoto et al. 35kV 33kV Electrode MCP EGA 130mm Ceramic bushing Voltage divider 100MΩ resister Cross section of V-IPM OTR Profile Monitor Y. Hashimoto et al. Gas-sheet Beam Profile Monitor Y. Hashimoto et al.

8 Beam Position Monitor (BPM) Same position as Q magnets (information of focus/defocus points) Diagonal cut electrode type (Main Ring) Parallel electrode type (Transport line) Diagonal cut type Quad parallel type Δ/Σ Δ/Σ Wire posi. Full aperture Wire position T. Toyama et al. D. Arakawa et al.

9 Main ring (MR) Injection Fast extraction (FX) Neutrino RF cavity Slow extraction (SX) Hadron Total length Energy 1568 m 3-30 GeV β Lorentz γ Harmonic 9 No. Bunches 8 Periods RF freq. Bunch length (time) Bunch length (space) µsec MHz 70~200 nsec 20~60 m Tune FX: ν x =22.40, ν y =20.75 SX: ν x =22.30, ν y =20.78 No. of BPM 186 (1 BPM/7-8 m) High intensity reduction of beam loss 500 m Stable beam orbit

10 Corrected COD Y. Sato Alignment errors are inevitable NEEDS COD correction CASE1 (2010 Aug) w USING steering magnets based on BPMs CASE2 (2011 Jun) 2.4 Accuracy of BPMs is a BIG KEY 0 Time (ms) 120 E13 protons per bunch 2.5 SCTR simulation for MR 150 kw eq. Model Alignment Simulated (DX-QcX) rms Simulated (DY-QcY) rms Simulated beam loss in injection for MR 150 kw eq. CASE1 (2010 Aug) 0.22 mm 0.19 mm 120 W (0.8%) CASE2 (2011 Jun) 0.42 mm 0.37 mm 220 W (1.5%) 10

11 J-PARC の Beam Position Monitor (BPM) Layout of electrode L R cosθ δl l δl U sinθ D θ T. Toyama

12 BPM system V L, V R, V U,V D : digitized signal (ADC out) Low Pass Filter Amp. Gain Different setup depending on beam intensity T. Toyama

13 Position calculation VV LL = λλgg LL VV RR = λλgg RR VV UU = λλgg UU 1 + xx aa 1 xx aa 1 + yy aa g L, g R, g U, g D Gains from electrode divided by Left gain (= g L ). g L =1 V L, V R, V U, V D Signal strength from electrode L,R,U,D λ Beam intensity passing through BPM x,y Beam positions (horizontal, vertical) a Effective radius from BPM center BPM alignment errors correction VV DD = λλgg DD 1 yy aa Beam Based Alignment (BBA) [2] xx = VV LL gg LL VV RR gg RR aa VV LL gg LL + gg RR VV RR yy = VV UU gg UU VV DD gg DD aa VV UU gg UU + gg DD VV DD (1) Most effective method to correct BPM position error very effective but takes long term (several days ~ 1 week) (2) Gain errors (signal transfer, electric circuit) Beam Based Gain Calibration (BBGC) [3] data acquisition is easier than BBA (a few~ several hours) [2] T. Toyama et al., PASJ meeting (2014). [3] K. Satoh and M. Tejima, Proc. of PAC95, p (1995).

14 BBGC (Beam Based Gain Calibration) 1) Kick the COD by steering magnet larger amplitude 2) Orbit data with various amplitudes are acquired. Signal from each electrode varies depending on beam positions (below fig.). 3) Gains (g L, g R, g U, g D ) are determined as reproducing the all signal strength from the electrodes. (-0.2,+0.2) y (mrad) (+0.2,+0.2) X : 5 points (x,y )=(ZSH001, ZSV216) Y : 5 points (ZSH210, ZSV209) X Y : 4 points Total14 points 14 points 10 shot 2 sets=280 data X (mrad) 2 phases pattern gains of BPM at nodes are not well reproduced. Amplitude (-0.2,-0.2) -0.2 (+0.2,-0.2) Beam direction -0.4

15 BBGC data analysis VV LL = λλgg LL 1 + xx aa Position xx = VV LL gg LL VV RR gg RR aa VV LL gg LL + gg RR VV RR yy = VV UU gg UU VV DD gg DD aa VV UU gg UU + gg DD VV DD VV RR = λλgg RR 1 xx aa Remove x,y,a λλ = 1 2 VV LL gg LL + VV RR gg RR Remove λ VV LL = 1 gg RR VV RR + 1 gg UU VV UU + 1 gg DD VV DD (gg LL = 1) VV UU = λλgg UU VV DD = λλgg DD 1 + yy aa 1 yy aa λλ = 1 2 VV UU gg UU + VV DD gg DD Simplified: RR gg RR + UU gg UU + DD gg DD = LL m equations, m:number of data RR 1 UU 1 DD 1 RR mm UU mm DD mm 1 ggrr 1 gguu = 1 ggdd LL 1 LL mm A x b Calculation of gains Solve the A x=b equations

16 Method (1) Least Square Fitting (LS) (2) Total Least Square Fitting (TLS) Minimize: residual R mm R = jj=1 RR jj gg RR + UU jj gg UU + DD jj gg DD LL jj 2 Minimize: total distance D DD = 1 GG mm 2 jj=1 GG = 1, 1 gg RR, RR jj gg RR + UU jj gg UU + DD jj gg DD LL jj 1 gg UU, 1 gg DD LL RR gg RR + UU gg UU + DD gg DD = 0 2 GG : plane equation in L,R,U,D: vector perpendicular to the plane

17 Method (1) Least Square Fitting (LS) (2) Total Least Square Fitting (TLS) Calculation procedure AA TT AA TT AA 11 AA TT AA xx = bb AA xx = AA TT bb AA xx = AA TT AA 11 AA TT bb xx = AA TT AA 11 AA TT bb AA TT AA xx = bb AA λλii xx = AA TT bb AA TT AA λλii 11 AA TT AA λλii xx = AA TT AA λλii 11 AA TT bb xx = AA TT AA λλii 11 AA TT bb λ: unknown const. I: Unit matrix

18 Simulation 1: Preparation of Gains VV LL = λλgg LL 1 + xx (g aa L, g R, g U, g D ) = (1.00, 1.01, 1.005, 0.975) 2: Determine positions VV LL = λλgg LL 1 + xx aa 2 x 2, 2 y 2, 25 points (right fig.) 3: Determine the signal for each position and gain VV LL = λλgg LL 1 + xx aa 4: Noise generation for V assuming V/V = 0.2% with Gaussian distribution 5: 500 data points are generated for 1 position λ: Coef. of beam int. a: calibrated value by offline

19 Simulation Conditions: True gains: (g L, g R, g U, g D ) = (1.00, 1.01, 1.005, 0.975) 2 x 2, 2 y 2, generated for 25 points Noise generation for V L, V R, V U, V D V/V = 0.2% Gaussian distribution 500 points per 1 position

20 Fitting results g R g U g D True LS TLS LS: Least Square Fitting TLS: Total Least Square Fitting # of data: m=25 500=12500 RR 1 UU 1 DD 1 RR mm UU mm DD mm 1 ggrr 1 gguu = 1 ggdd LL 1 LL mm TLS method can adequately reproduce the true gains.

21 Analysis (Beam data) Gains vary depending on the settings of the circuit Proton/8-bunch Amp. gain Low Pass Filter Low Int OFF High Int ON Low Pass Filter Amp. Gain Signal from electrode Low Int. High Int. Waveform from each electrode FFT spectra of waveform Low Int MHz 5.01 MHz High Int MHz (used for analysis) Time (µsec) Gains are calculated for 3.34 MHz peak (signal strength from L,R,U,D) Freqency (MHz)

22 Results of gain calculation Low Int. High Int. g R /g L Maximally a few-several % difference g U /g L BBGC is needed depending on beam intensity g D /g L BPM address No.

23 Difference of gains by derived by different Runs Low Int. 14/04/01 14/11/29-30 High Int. 14/11/ /11/29-30 g D /g L g U /g L g R /g L BPM address No. BPM address No.

24 Difference of gains by derived by different Runs Low Int. High Int. g D /g L g U /g L g R /g L BPM address No. BPM address No.

25 Evaluation of the results Evaluated by Root Mean Square (RMS)= Low Int. xx2 ii nn of COD position for Low and High beam intensities High Int. x (mm) black: ref. gain, blue: cal. gain black: ref. gain, red: calc. gain y (mm) BPM addr. # BPM addr. # Low Int. High Int. RMS x RMS y RMS x RMS y Ref. gain Cal. gain

26 Summary (mainly Beam Based Gain Calibration) 186 BPMs are used in J-PARC MR for COD correction Required accuracy ~ a few 100 µm Correction of alignment errors of BPMs is necessary BPM gain has individuality by signal transfer or electric circuit. Beam Based Gain Calibration (BBGC) is effective method to correct position error along with Beam Based Alignment (BBA) Gains vary with the setting of electric circuit depending on beam intensity. BBGC has been done for Low and High beam intensity. RMS of COD was improved for x position while RMS of y became worse under investigation (real difference in gains or some errors?) Establishment of BBGC for various beam intensities and will be applied for corrections of position errors.