J.J. García-Garrigós Septiembre 2008

J.J. García-GarrigósSeptiembre 2008

126/09/2008

Design and Construction of a Beam Position Monitor Prototype

for the Test Beam Line of the CTF3

226/09/2008

Contents

Introduction: Linear Collliders

The CLIC and CTF3

The BPS monitor prototype in theTest Beam Line

BPS mechanical design

BPS sensing mechanism and general description

BPS electronic design

BPS wire test results and analysis

Conclusions and Future work

BPS-TBL-CTF3 J.J. García-Garrigós

3

The LHC will probe the new “terascale” energy region : • Confirm or refute the existence of the Higgs boson to

complete the Standard Model• Explore the possibilities for physics beyond the

Standard Model, such as supersymmetry, extra dimensions and new gauge bosons

26/09/2008 BPS-TBL-CTF3 J.J. García-Garrigós

Present and Future Colliders

Particle physics community worldwide have reached a consensus that the results from the LHC will need to be complemented by experiments at an electron-positron collider operating in the tera-electron-volt energy range

4

• p-p colliders can reach higher energy than e+e-, but– the energy of the constituents (quarks and gluons) are lower– p-p interaction is too complicated (not easy to analyze collision data)

• e+e- colliders:– cleaner experimental enviroments– available eγ, γγ and e-e- interactions– available polarized beams

• p-p and e+e- complementary– particle discovery by p-p colliders – finer study by e+e- colliders Naturally arise as LHC successors


Why e+ e- Linear Colliders

Some physics reasons


5

LEP at CERN, CHEcm = 180 GeVPRF = 30 MW

Why a Linear Collider,and not just build a bigger Storage Ring

500 GeV LC

Livingston Chart


Why a Linear Collider,and not just build a bigger Storage Ring

2222

2BECceP

4

/EC

revE

B22

22

2BECceP

Synchrotron radiation from an e- in a magnetic field:

Energy loss per turn of a machine with an average radius :

Energy loss per turn has to be be replaced by the RF system, which is the major cost factor for a collider.

ecEB

average power

LEP e+e- storage ring:The biggest superconducting RF system

with3640 MV per revolutionjust enough to keep the beam in LEP at its nominal energy

7

e+ e-

5-10 km

No Synchrotron Radiation, but new problems arise:• we cannot store the beams, LC is one-pass device where the beams

must be accelerated to the required energy on each pulse of the machine

• we cannot take advantage of the stored beam to slowly ramp the energy up, we have to provide several Km of RF Power (25-100 MV/m) to achieve the energy in a single-pass


Why a Linear Collider,cause no bends, but also needs lots of RF!

8

This study is based on an RF system using superconducting cavities for acceleration, with a nominal accelerating field of 31.5 MV/m and a total length of 31 km for a colliding-beam energy of 500 GeV.

The CLIC scheme is based on normal conducting travelling-wave accelerating structures, operating at a frequency of 12 GHz and with very high electric fields of 100 MV/m to keep the total length to about 48 km for a colliding-beam energy of 3 TeV.

¿ ?


Now, no losses And we can get the RF Power

Feasible

Future Linear Colliders

Proof of Principle

9

The peak RF power required to reach the electric fields of 100 MV/m amounts to about 275 MW per active meter of accelerating structure. Not possible with klystrons.


CLIC: The Compact LInear Collider

Each sub-system pushes the state-of-the art in accelerator design

Hence a novel power source, an innovative two-beam acceleration system, in which another beam, the drive beam, supplies energy to the main accelerating beam.

10

• To demonstrate the Two-beam

acceleration scheme.

• A scaled facility for one branch of the

Drive Beam Generation System

Layout of the CLIC EXperimental area (CLEX) building with TBL


CTF3: The CLIC Test Facility 3

11

16 TBL Cells


TBL: The Test Beam Line

The main aims of the TBL: Study and demonstrate the technical feasibility

and the operability a drive beam decelerator (including beam losses), with the extraction of as much beam energy as possible. Producing the technology of power generation needed

for the two-beam acceleration scheme.

Demonstrate the stability of the decelerated beam and the produced RF power by the PETS.

Benchmark the simulation tools in order to validate the corresponding systems in the CLIC nominal scheme.

12

TBL + BPM specifications Main features of the Inductive Pick-Up (IPU) type of BPM:• less perturbed by the high losses experienced

in linacs;• the total length can be short; • it generates high output voltages for typical

beam currents in the range of amperes; • calibration wire inputs allow testing with

current once installed• Broadband, but better for bunched beams with

short bunch duration or pulse IPU type of BPM suitable for TBL


2 BPS Prototypes developed at IFIC,scaled and redesigned version of IPU used in DBL of CTF3

TBL beam time structure

13

BPS Mechanical Assembly


Vacuum assembly: ceramic tube with Kovar collars at both ends, one collar TIG welded to the downstream flange, and the other one electron welded to a bellow and a rotatable flange. ~10-10 mbarl/s [High Vacuum]

Ferrite cylinder

Cooper body

PCB plates

14

BPS Basic Sensing Mechanism

Four Outputs with two Calibration inputs: [V+,V-, H+,H-] and [Cal+, Cal-], respectively

Difference signals (Δ) normalized to sum signal (Σ) proportional to beam position coordinate,

xV α ΔV /Σ [Vertical plane] xH α ΔH /Σ [Horizontal plane]where: ΔV ≡ (V+ − V-); ΔH ≡ (H+ − H−);and, Σ ≡ (V+ + H+ + V− + H- ) 26/09/2008 BPS-TBL-CTF3

J.J. García-Garrigós

Primary transformer electrode

Longitudinal cross-section view

15

BPS Readout chain


Amplifier developed at UPC by G. Montoro

Digitizer/ADC developed at LAPP(Annecy)

Both Designs must be Rad-Hard

16

Typical IPU Frequency Response


Induced current/signal Pulse deformation

ωlow = R/L, and ωhigh = 1/RCS

τdroop =1/ ωlow , and τrise =1/ωhigh

τdroop ~ 102 tpulse τrise ~ 10-2 tpulse

To let pass the pulse without deformation

Droop time very important for ADC sampling.

17

BPS Electronic design

Characteristic Output Signal Levels:


with: (Σ /IB) = 0.55ΩVsec = (RLoadRS1/(RS1+RS2+RS1)N) Ielec ≡ (Σ/IB) Ielec

PCBs Schematics and Output relation

For a beam current of: IB = 30AΣ = 16.5 V [outputs sum]Vsec = Σ /4 = 4.125V [centered beam] ||ΔV||max = ||ΔH||max = Σ /2 = 8.25V [beam at elec]

for the design values: RLoad = 50 Ω, RS1 = 33 Ω, RS2 = 18 Ω and N = 30 turns

18

BPS1 Characterization Tests [The Wire-Test]*

Sensitivity and Linearity+

Frequency Response


carried out during several short stays at CERN, in the AB/BI-PI[1], where the wire testbench is placed, and it has been previously used for testing and calibrating BPMs for the Drive Beam Linac (DBL) of the CTF3.

Tests carried out during several short stays at CERN, in the AB/BI-PI* Labs (Bldg.37)

Testbench used to characterize the BPMs for the Drive Beam Linac (DBL) of the CTF3

*With the help of: CTF3 Collaboration

* Accelerator an Beams Department/ Beam Instrumentation Group – Position and Intensity Section

19

Sensitivity and Linearity Test ResultsSensitivity


Electric Offset

Sensitivity for V,H planes Electric Offset for V,H planes

Linear fit equations

SV = (41.09±0.08)10−3 mm−1

SH = (41.53±0.17)10−3 mm−1 EOSH = (0.15±0.02) mm

EOSV = (0.03±0.01) mm


Sensitivity and Linearity Test ResultsLinearity errorOverall Precision/Accuracy

σV = 78 μm

σH = 170 μm

ii) Misalignment in the horizontal electrodes i) Low current in the wire (13mA) vs beam 32 AσTBL < 50μm

Typical S-shape

BPS above specs

21

Cut-off Frequencies:


Frequency ResponseTest ResultsOutput electrodes ΔV, ΔH and Σ

Wire

Pos

: Cen

ter

Wire

Pos

:+8m

m V

,H

fLΣ = 1.76 KHz fLΔ ≡ fLΔH = fLΔV = 282KHz τdroop Σ = 90us τdroop Δ = 564ns

Bandwidth specs:[10KHz-100MHz] tpulse=140ns fhigh > 100 MHz, and τrise < 1.6 ns

Couplinglow freq. components don’t feel the beam variation

22

BPS Electric Model

Low cut-off frequenciesTwo different cases:I) Centered wire: Balanced wall image curentII) Displaced wire: Unbalanced wall image current (low freq. coupling)


High cut-off frequencyFixed by secondary Cs for all cases

fhigh = 1/2 R𝜋 eCS

Model Cut-off Frequencies:

23

BPS Electric Model

I) Centered wire: Balanced wall image curent:• Δ ~0 LΔ = 0 because reflects a coupling in the other case• Low cut-off fixed by LΣ >>LΔ f Σ << fΔ


24

BPS Electric Model


II) Displaced wire V,H plane: Unbalanced wall image current (low freq. coupling)• Δ ≠0 LΔ ≠0 appears on the pair of V or H electrodes • Low cut-off fixed by LΔ >> LΣ fΔ general case and must be

compensated by External Amplifier

25

BPS Electric Model

II) Displaced wire V,H plane: Unbalanced wall image current (low freq. coupling)• Δ ≠0 LΔ ≠0 appears on the pair of V or H electrodes • Low cut-off fixed by LΔ >> LΣ fΔ general case and must be

compensated by External Amplifier



Pulse Response and Calibration problemfLΔ[cal] =180 KHz < fLΔ =282 KHz their difference is about 100 KHz.

Represents a problem for the amplifier compensation in the Δ channels, to lower the Δ low cut-off frequency for the wire, fLΔ; because the same compensation designed for the fLΔ will be applied when exciting the calibration inputs to fLΔ[Cal]Bad Pulse for calibration (overcompensation).

A compromise solution: compensation frequency at the lower one, fLΔ[Cal] Cal. pulse good flatness and wire-beam pulse flat enough for TBL pulse duration(140ns)

τdroop Δ [cal] = 884 ns

τdroop Δ = 564 ns

τdroop Σ = 90 μs

τdroop Σ [Cal] = 90 μs

27

Conclusions and Future Work


A set of two BPS prototypes with the associated electronics were designed and constructed.

The performed tests yield:

• Good linearity results and reasonably low electrical offsets from the mechanical center.

• Good overall-precision/accuracy in the vertical plane considering the low test current; and, a misalignement in the horizontal plane was detected by accuracy offset and sensitivity shift.

• Low frequency cut-off for Σ/electrodes signals, fLΣ, and high cut-off frequency, fhigh, under specifications.

• Low frequency cut-off for Δ signals, fLΔ, determined to perform the compensation of droop time constant, τdroopΔ, with the external amplifier.

28



Open issues for improvement in the BPS2 monitor prototype:• Correct the possible misalignments of the horizontal plane electrodes

suggested in the linearity error analysis.• Check if overall-precision below 50μm (under TBL specs), with enough

wire current New wire testbench at IFIC. • Study the different low cut-off frequencies in the calibration, fLΔ[Cal], and

wire excitation cases, fLΔ.

Test Beam of the BPS1 in the TBLResolution at maximum current.

BPS’ Series production and characterization (15 more units). The new wire testbench will allow accurate (anti-vibration + micro-movement system) and automatized measurements.

29



Sketch of New IFIC Wire Testbench. Under development right now.


Thanks for your Attention

Muchas Gracias


BPS1 Characterization Table BPS1 Sensitivity and Linearity Parameters

Vertical Sensitivity, SV 41.09 mm-1

Horizontal Sensitivity, SH 41.43 mm-1

Vertical Electric Offset, EOSV 0.03 mmHorizontal Electric Offset, EOSH 0.15 mm

Vertical overall precision (accuracy), σV 78 μm

Horizontal overall precision (accuracy), σH 170 μmBPS1 Characteristic Output Levels

Sum signal level, Σ 16.5 VDifference signals max. levels, ||ΔV||max, ||ΔH||max 8.25 V

Centered beam level, Vsec (xV = 0, xH = 0) 4.125 VBPS1 Frequency Response (Bandwidth) Parameters

Σ low cut-off frequency, fLΣ 1.76 KHzΔ low cut-off frequency, fLΔ 282 KHzΣ low cut-off frequency calibration, fLΣ [Cal] 1.76 HzΔ low cut-off frequency calibration, fLΔ [Cal] 180 KHzHigh cut-off frequency, fhigh > 100 MHzHigh cut-off frequency calibration, fhigh [Cal] > 100 MHz

BPS1 Pulse-Time Response ParametersΣ droop time constant, τdroopΣ 90 μs

Δ droop time constant, τdroopΔ 564 ns

Σ droop time constant calibration, τdroopΣ [Cal] 90 μs

Δ droop time constant calibration, τdroopΔ [Cal] 884 μsRise time constant calibration, τrise < 1.6 nsRise time constant calibration, τrise [Cal] < 1.6 ns


BPS Monitors Schedule

J.J. García-Garrigós Septiembre 2008

Documents

collidingbeam energy

energy loss

collision datae e colliders

butthe energy

higher energy

required energy

garcagarrigs present

pp colliders finer study