Muon (g-2) Calorimeter Calibration System Carlo Ferrari

MidTerm Review, Pisa, March 4-5, 2019

NEWS NEw WindowS on the universe and technological advancements from trilateral EU-US-Japan collaboration

Web site: risenews.df.unipi.it

2

•  ThemaintaskofthelasersystemintheMuonexperiments(bothG-2andMu2e)isthecorrec=onofthegaindri?andinstabilityofthephotodetectors(SiPM)ofthecalorimeters(G-2:54SiPMseachfor24calorimeters,Mu2e:900SiPMseachfor2calorimeters),whichareamajorcauseofsystema=cerror.

•  ThemainfactorsthatinfluencetheSiPMgainisthetemperature(gaindri?);however,intheG-2experiment,thepar=clesplashintheearlystageofthemuonfillproduceagainsagging,duetohighpar=clesrate.

Introduc=on

3

Secondments:•  13thFeb2018–26thFeb2018•  19thJun2018–13thAug2018DuringthefirstsecondmentIworkedonthereduc=onofgainsagging,addingcapacitorsneartheSiPMs(embeddingitintheso-calleddogbox).Totestthesolu=on,alasermodewasimplementedthatprovidesasimulatedsplash(100laserpulseswithrepe==onrateofabout8MHz)followedbyprobepulsesspacedtemporallyby5us.Thelaserpulsessentatthebeginningofthefillwereeffec=veinsimula=ngthesplash.Thetestsdemonstratedthattheaddedcapacitorswereeffec=veinfla\eningthegaincurveoftheSiPMsandbiggercapacitors(megabox)havebeeninstalledoneachcalorimeter.

Ferrari’ssecondements

Gainfunc=onduringthemuonfill,forthe54channelsofcalorimeter1:(a)before(run9183)and(b)a?er(run13598)themegaboxinstalla=on(thankstoAnnaDriu^).

(a) (b)

4

Duringthesecondsecondment(19thJun2018–13thAug2018)Iworkedonthefollowingtasks:§  Ihavedone6shi?s(8hourseach)inthecontrolroomoftheG-2experiment(June)

§  Ia\endedtheMu2eCollabora=onMee=ng(Jun26thtoJun29th,2018)

SecondFerrari’ssecondement

SM

Laser1

Collimator

DiffuserFiberharps3and4

Cube90:10Cube70:30

Filterwheel

Launchingfiber1

Launchfiber2Launchfiber3

Launchfiber4

Cube50:50

DiffuserFiberharps1and2

Collimator

Filter80%Toberemoved

20msilicafibers

Fiberbundletotheharps

Fiberbundle

§ Idesignedandinstalledop=calcomponentsonthelaserhutinordertoprovidelaserspulsestothetwoFiberHarpdetectorsinthering.Lasern.1hasbeensampledwitha90:10cube,thesampledbeamhasbeenspli\edintwobeamswitha50:50cube,thetwobeamscollimatedintotwo20mlongSilicafibers.Attheendofthetwofibersthereareop=calcomponentsinordertoobtainaflat-topbeam,½”wide.Attheoutputofeachop=calsystemstherearetwofibersbundles(32fiberseach)providingthelaserpulsestotheSiPMsembeddedinthetwoFiberHarpdetectors.Weinstalledthetwoop=calfibers,checkedthetransmissionoftheop=calsystem.Furthermore,afilterwheel(12filter)hasbeeninstalled,inordertochangetheamplitudeofthelaserpulses.

G-2,Aug2018:LaserpulsestotheFiberHarps

5

Installa=onoftheop=csinthelaserhut,oftwonewop=calfibersandtwofan-outsystems

16fibersbundle

Engineereddiffuser

6

Furthermore,Iproposed,designedandIhavebeeninvolvedinthestudyofthesynchronisa=onofthelaserpulsessenttothe24calorimeters.EachlaunchingfiberconnectedtothecalorimetershasbeendisconnectedandconnectedtoaPMTalongwithanotherfiberthatcarriesareferencesignalfromanothercalorimeter.Thedifferencebetweenthearrival=meofthetwopulsesprovidesthe=medelaybetweenthelaserpulses,duetothelaserheadsandthelengthofthecablesthatcarrythetriggersignalstothecontrolboardsofthesixlasers.Wemeasuredthe=medifferencewithincalorimetrswithaccuracyof100ps(wellbelowtherequested500pssynchronisa=on).

SecondFerrari’ssecondement

1

24

23

Laserhut

PMT20mfiberoscilloscope

1mfiber

Calorimeters

Togetthehardwaredelaybetweencalorimeterswesendtwopulsesfromtwocalorimetersthroughtwofibersofdifferentlength.

G-2laserpulsessynchronisa=on

7

1

2

3

4

5

6 4

3

2

1

laserheads

lauchingfibers

calorimeters(diffuser)

54crystals ridersboards

lasercontrolboard

δF≈0-2ns

TF

δL≈0.4ns

TR(fill)TL

δR≈0-3ns

ResultsonG-2hardwaremeasurementofδF

8(ThankstoPaoloGiro^)

9

Furthermore,Imadefewtestregardingthelasercalibra=onsystemfortheMu2eexperiment.Theexperimentalsetuptosplitthelaserinto8beamshasbeeninstalledandsuccessfullyalignedinthelaserroomatSiDet,using750:50spli^ngcubesinacompactdesign(20cmx20cm).Thetransmissionofthesystemisabout8%foreachbeam,theuniformityofintensityiswithin15%.

SecondFerrari’ssecondement

9

Laser

Fiber collimator

Filter wheel

Monitor

Cube splitting

Mu2eLasersystemscheme

10

ToDisk1

ToDisk2

Photodiode

Photodiode

Photodiode

LaserVacuum

Filterwheel Monitor

Collimator

Mu2e Technical Design Report

Fermi National Accelerator Laboratory

9-56

Figure 9.61. Picture of the ThorLab IS-200 integrating sphere (left); and the sphere’s reflectivity dependence on wavelength.

There is not a stringent requirement on the laser pulse width, since the APD readout electronics has a rise time between 6 to 8 ns, thus setting an upper limit on the width of 10 ns. Similarly, the pulse frequency is not strongly constrained since, as shown in the prototype test, running at 1 Hz provides better than per-mil statistical precision in one hour of data-taking. It is instead mandatory to synchronize the laser pulse with an external trigger to allow the light to reach the detector at the correct time relative to the proton beam pulse so that laser data can be taken during the time when the calorimeter is acquiring physics data as well as during the gaps between beam when the calorimeter is quiet. The laser pulse energy is strongly attenuated by the distribution system. However, the laser signal is required to simulate a 100 MeV energy deposition. For BaF2 this corresponds to ~10,000 p.e. in each photosensor. This roughly translates to a 10-20 nJ energy source. A safety factor of 20 is designed into the system to account for the eventual degradation of the signal transmission with time, resulting in an energy pulse requirement of ~ 0.5 µJ. There is a stringent requirement on the fibers. They should have high transmission at 200-260 nm, a small attenuation coefficient and they must be radiation hard up to O(100 krad). The best choice is fused silica fibers, both for their transmission properties (see Figure 9.62, left), a nearly flat wavelength dependence down to 150 nm, a long attenuation length and high radiation tolerance.

9.12.1 Laser monitor prototype for the LYSO crystals The setup used for the transmission test and for the calibration of the LYSO calorimeter prototype is shown in Figure 9.62 (right). The light source was an STA-01 solid-state pulsed laser emitting at 532 nm with a pulse energy of 0.5 µJ, a pulse width < 1 ns, good pulse-to-pulse stability (3%), and synchronization to an external trigger for frequencies up to 100 kHz. Table 9.7 summarizes the performance of equivalent STA-01 lasers

Mu2e Technical Design Report

Fermi National Accelerator Laboratory

9-56

Figure 9.61. Picture of the ThorLab IS-200 integrating sphere (left); and the sphere’s reflectivity dependence on wavelength.

There is not a stringent requirement on the laser pulse width, since the APD readout electronics has a rise time between 6 to 8 ns, thus setting an upper limit on the width of 10 ns. Similarly, the pulse frequency is not strongly constrained since, as shown in the prototype test, running at 1 Hz provides better than per-mil statistical precision in one hour of data-taking. It is instead mandatory to synchronize the laser pulse with an external trigger to allow the light to reach the detector at the correct time relative to the proton beam pulse so that laser data can be taken during the time when the calorimeter is acquiring physics data as well as during the gaps between beam when the calorimeter is quiet. The laser pulse energy is strongly attenuated by the distribution system. However, the laser signal is required to simulate a 100 MeV energy deposition. For BaF2 this corresponds to ~10,000 p.e. in each photosensor. This roughly translates to a 10-20 nJ energy source. A safety factor of 20 is designed into the system to account for the eventual degradation of the signal transmission with time, resulting in an energy pulse requirement of ~ 0.5 µJ. There is a stringent requirement on the fibers. They should have high transmission at 200-260 nm, a small attenuation coefficient and they must be radiation hard up to O(100 krad). The best choice is fused silica fibers, both for their transmission properties (see Figure 9.62, left), a nearly flat wavelength dependence down to 150 nm, a long attenuation length and high radiation tolerance.

9.12.1 Laser monitor prototype for the LYSO crystals The setup used for the transmission test and for the calibration of the LYSO calorimeter prototype is shown in Figure 9.62 (right). The light source was an STA-01 solid-state pulsed laser emitting at 532 nm with a pulse energy of 0.5 µJ, a pulse width < 1 ns, good pulse-to-pulse stability (3%), and synchronization to an external trigger for frequencies up to 100 kHz. Table 9.7 summarizes the performance of equivalent STA-01 lasers

Mu2e Technical Design Report

Fermi National Accelerator Laboratory

9-56

Figure 9.61. Picture of the ThorLab IS-200 integrating sphere (left); and the sphere’s reflectivity dependence on wavelength.

There is not a stringent requirement on the laser pulse width, since the APD readout electronics has a rise time between 6 to 8 ns, thus setting an upper limit on the width of 10 ns. Similarly, the pulse frequency is not strongly constrained since, as shown in the prototype test, running at 1 Hz provides better than per-mil statistical precision in one hour of data-taking. It is instead mandatory to synchronize the laser pulse with an external trigger to allow the light to reach the detector at the correct time relative to the proton beam pulse so that laser data can be taken during the time when the calorimeter is acquiring physics data as well as during the gaps between beam when the calorimeter is quiet. The laser pulse energy is strongly attenuated by the distribution system. However, the laser signal is required to simulate a 100 MeV energy deposition. For BaF2 this corresponds to ~10,000 p.e. in each photosensor. This roughly translates to a 10-20 nJ energy source. A safety factor of 20 is designed into the system to account for the eventual degradation of the signal transmission with time, resulting in an energy pulse requirement of ~ 0.5 µJ. There is a stringent requirement on the fibers. They should have high transmission at 200-260 nm, a small attenuation coefficient and they must be radiation hard up to O(100 krad). The best choice is fused silica fibers, both for their transmission properties (see Figure 9.62, left), a nearly flat wavelength dependence down to 150 nm, a long attenuation length and high radiation tolerance.

9.12.1 Laser monitor prototype for the LYSO crystals The setup used for the transmission test and for the calibration of the LYSO calorimeter prototype is shown in Figure 9.62 (right). The light source was an STA-01 solid-state pulsed laser emitting at 532 nm with a pulse energy of 0.5 µJ, a pulse width < 1 ns, good pulse-to-pulse stability (3%), and synchronization to an external trigger for frequencies up to 100 kHz. Table 9.7 summarizes the performance of equivalent STA-01 lasers

Mu2e Technical Design Report

Fermi National Accelerator Laboratory

9-56

Figure 9.61. Picture of the ThorLab IS-200 integrating sphere (left); and the sphere’s reflectivity dependence on wavelength.

There is not a stringent requirement on the laser pulse width, since the APD readout electronics has a rise time between 6 to 8 ns, thus setting an upper limit on the width of 10 ns. Similarly, the pulse frequency is not strongly constrained since, as shown in the prototype test, running at 1 Hz provides better than per-mil statistical precision in one hour of data-taking. It is instead mandatory to synchronize the laser pulse with an external trigger to allow the light to reach the detector at the correct time relative to the proton beam pulse so that laser data can be taken during the time when the calorimeter is acquiring physics data as well as during the gaps between beam when the calorimeter is quiet. The laser pulse energy is strongly attenuated by the distribution system. However, the laser signal is required to simulate a 100 MeV energy deposition. For BaF2 this corresponds to ~10,000 p.e. in each photosensor. This roughly translates to a 10-20 nJ energy source. A safety factor of 20 is designed into the system to account for the eventual degradation of the signal transmission with time, resulting in an energy pulse requirement of ~ 0.5 µJ. There is a stringent requirement on the fibers. They should have high transmission at 200-260 nm, a small attenuation coefficient and they must be radiation hard up to O(100 krad). The best choice is fused silica fibers, both for their transmission properties (see Figure 9.62, left), a nearly flat wavelength dependence down to 150 nm, a long attenuation length and high radiation tolerance.

9.12.1 Laser monitor prototype for the LYSO crystals The setup used for the transmission test and for the calibration of the LYSO calorimeter prototype is shown in Figure 9.62 (right). The light source was an STA-01 solid-state pulsed laser emitting at 532 nm with a pulse energy of 0.5 µJ, a pulse width < 1 ns, good pulse-to-pulse stability (3%), and synchronization to an external trigger for frequencies up to 100 kHz. Table 9.7 summarizes the performance of equivalent STA-01 lasers

Thelaserpulsesaresentsimultaneouslyto1200crytals/SiPMslocatedin2calorimetersinthevacuumchamberthrough860mlongfibersand8integra=ngspheres.24fibersbundlesperformthesecondfan-out.

Primarydistribu=on

Secondarydistribu=on

Conclusions

11

•  TheG-2lasersystemhasbeeninstalledandisproperlyworking

•  TheMu2elasersystemhasbeendesigned•  TheMu2eprimarydistribu=onsystemhasbeenassembled,

alignedandtested