- Email: [email protected]
Results from the FDIRC prototype
Nuclear Instruments and Methods in Physics Research A ∎ (∎∎∎∎) ∎∎∎–∎∎∎
1 2 3 4 5 6 7 8 9 10 11 12 13 14 Q1 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66
Contents lists available at ScienceDirect
Nuclear Instruments and Methods in Physics Research A journal homepage: www.elsevier.com/locate/nima
Results from the FDIRC prototype D.A. Roberts a,n, N. Arnaud b, B. Dey c, M. Borsato b, D.W.G.S. Leith d, K. Nishimura d, B.N. Ratcliff d, G. Varner e, J. Va’vra d a
University of Maryland, College Park, MD 20742, USA Laboratoire de l’Accélérateur Linéaire, Centre Scientifique d’Orsay, F-91898 Orsay Cedex, France c University of California, Riverside, CA 92521, USA d SLAC, Stanford University, Palo Alto, CA 94309, USA e University of Hawaii, Honolulu, HI 96822, USA b
art ic l e i nf o
Keywords: Cherenkov detectors RICH Photodetectors B factory
a b s t r a c t We present results from a novel Cherenkov imaging detector called the Focusing DIRC (FDIRC). This detector was designed as a prototype of the particle identification system for the SuperB experiment, and comprises 1/12 of the SuperB barrel azimuthal coverage with partial electronics implementation. The prototype was tested in the SLAC Cosmic Ray Telescope (CRT) which provides 3-D muon tracking with an angular resolution of 1.5 mrad, track position resolution of 5–6 mm, start time resolution of 70 ps, and a muon low-energy cutoff of 2 GeV provided by an iron range stack. The quartz focusing photon camera couples to a full-size BaBar DIRC bar box and is read out by 12 Hamamatsu H8500 MaPMTs providing 768 pixels. We used IRS2 waveform digitizing electronics to read out the MaPMTs. We present several results from our on-going development activities that demonstrate that the new optics design works very well, including: (a) single photon Cherenkov angle resolutions with and without chromatic corrections, (b) S/N ratio between the Cherenkov peak and background, which consists primarily of ambiguities in possible photon paths to a given pixel, (c) dTOP ¼TOPmeasured– TOPexpected resolutions, and (d) performance of the detector in the presence of high-rate backgrounds. We also describe data analysis methods and point out limits of the present performance. & 2014 Elsevier B.V. All rights reserved.
1. Introduction The DIRC detector at the BaBar experiment provided excellent particle identification performance [1]. Based on this success, we have been pursuing an R&D program to develop a compact and fast detector for future particle identification systems. One such idea, a Focusing DIRC (FDIRC) [2], would be capable not only of measuring the (x,y) coordinate of each photon with an angular resolution similar to the BaBar DIRC, but would also measure each photon's time-of-propagation (TOP) along the fused silica bar with 100–200 ps single-photoelectron timing resolution (the BaBar DIRC had a timing resolution of s 1.6 ns). The precise timing would allow one to correct for the chromatic dispersion contribution to the Cherenkov angle resolution and thus improve the angle measurement substantially ( 1 mrad). The improved timing resolution would also provide better background rejection, an important feature for detectors operating in the highintensity environment of a Super Flavor Factory for example, and
n
Corresponding author. Tel.: þ 1 301 405 6067; fax: þ 1 301 314 9896. E-mail address: [email protected] (D.A. Roberts).
could also provide modest improvement to the particle separation in some regions of phase space through the particle time of flight. By using photon detectors with a small pixel size one can reduce the size of the expansion volume by up to a factor of 10 relative to the BaBar DIRC while maintaining similar spatial resolution. The smaller geometric size together with better timing resolution would improve background suppression by nearly two orders of magnitude. The first prototype of the FDIRC concept was constructed and operated in a test beam in 2005, 2006, and 2007 [3,4]. This prototype utilized a single DIRC quartz bar and a cylindrical mirror placed in a mineral oil expansion volume. It was the first RICH detector to successfully correct the chromatic error using timing. Here, we describe a FDIRC design developed for the SuperB experiment that was to be built in Fascati [5,6]. Fig. 1 shows a GEANT4 realized image of the FDIRC, indicating key components.
2. Description of the FDIRC prototype The particle identification system for the SuperB experiment, or any other high-intensity Super Flavor Factory experiment, must
http://dx.doi.org/10.1016/j.nima.2014.04.069 0168-9002/& 2014 Elsevier B.V. All rights reserved.
Please cite this article as: D.A. Roberts, et al., Nuclear Instruments & Methods in Physics Research A (2014), http://dx.doi.org/10.1016/j. nima.2014.04.069i
2
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66
D.A. Roberts et al. / Nuclear Instruments and Methods in Physics Research A ∎ (∎∎∎∎) ∎∎∎–∎∎∎
utilize 1.5 μs of its 12 μs analog storage buffer. The digitized waveforms were analyzed off-line. The entire assembly was placed in the SLAC Cosmic Ray Telescope (CRT) [8]. The telescope features include: two planes of scintillator hodoscopes that provide 3-D tracking of cosmic ray muons with an angular resolution of 1.5 mrad and track position resolution of 5–6 mm; a quartz start counter providing start time resolution better than 70 ps; and an instrumented iron range stack used to select muons with energy greater than 2 GeV. The scintillator hodoscopes had active areas of 51 cm 107 cm and provided an angular tracking acceptance of 7 171 7271 in two respective directions. The bar box was located within the CRT such that the cosmic muons were incident near the midpoint along the length of the quartz radiators. The results to follow are based upon over 600,000 cosmic ray triggers in the CRT.
3. Experimental results
Fig. 1. The FDIRC optics as simulated in GEANT4.
cope with much higher luminosity related background rates than in the previous generation's experiments such as BaBar while maintaining similar physics performance. The estimated increase in background rates is on the order of 100-times higher. The basic design strategy for dealing with these increased rates was to make the camera much smaller and much faster. The design decision was to use a new photon camera-imaging concept based on focusing optics. The focusing element and expansion volume responsible for imaging the Cherenkov photons on to the PMT cathode surfaces were machined from radiation hard pieces of fused silica. In addition to the reduced background sensitivity gained from the geometric reduction in size, the use of fused silica also reduces background sensitivity, especially to neutrons, compared to a water-filled expansion volume. The improved timing not only helps with background suppression, but can also be used to measure and correct for the chromatic dispersion, thus improving performance. By including a cylindrical mirror that focuses light on to the photodetector plane we can remove the pinholesize (i.e. bar thickness) component of the angular resolution in one dimension of the quartz bar. The prototype was constructed utilizing a spare bar box from the BaBar DIRC. This is a complete bar box with 12 quartz bar radiators. The individual quartz bars had nominal dimensions of 35 mm wide 17.25 mm high 4900 mm long. The new optical components are a new wedge and a large focusing block (FBLOCK), both manufactured from solid fused silica [6]. The dimensions of the FBLOCK were nominally 560 mm high 217 mm long 422 mm wide. The new wedge was coupled to the bar box window with 50–75 mm thick epoxy. The FBLOCK was then coupled to the new wedge with 1 mm thick RTV. The entire assembly was housed in a light-tight mechanical support structure that connects to the bar box. The focal plane of the FBLOCK was populated with 12 Hamamatsu H8500 multi-anode photomultiplier tubes (MaPMT). This only partially covers the focal plane of the FBLOCK; complete coverage would require 48 MaPMTs. The location of the 12 MaPMTs was chosen to optimize the coverage for the expected hit locations from cosmic ray muons. Each MaPMT provides 64 pixels, with each pixel being approximately 6 6 mm2. The MaPMTs were read out with IRS2 [7] waveform digitizing electronics, preceded by a 40 preamplifier. The IRS2 was operated at a sampling rate of 2.7 GHz, and was configured to
To reconstruct the Cherenkov angle in a given event, we used a Monte Carlo generated dictionary that maps MaPMT pixel to photon direction inside the quartz bar, kpixel , as well as the time of propagation of the photon inside the FBLOCK. The dictionary was generated by simulating single photons in the quartz, drawn from an isotropic initial angular distribution. There may be more than one photon path that will lead from the quartz bar to a given pixel, giving rise to multiple solutions of kpixel for a given MaPMT hit. The dictionary is therefore multi-valued. Many of these ambiguous solutions can be eliminated either via timing information or because they give rise to unphysical values of the Cherenkov angle. The cosine of the Cherenkov angle was then calculated as the dot product of the track direction and the photon direction: cos θC ¼ ktrack Ukpixel [9]. The expected time-of-propagation (TOP) of a photon can be calculated from the track's entry location in the quartz bar as measured by the CRT, the photon direction kpixel , and the time-ofpropagation inside the FBLOCK taken from the Monte Carlo dictionary. This was then compared to the measured TOP, dTOP ¼ TOP measured TOP expected . This variable, dTOP, provides the primary background rejection and ambiguity resolving power. Fig. 2 shows the measured dTOP for forward traveling (direct) and backward traveling (indirect) photons. The difference in dTOP resolution between forward and backward photons is primarily explained by chromatic dispersion and is well modeled by the Monte Carlo simulation. Fig. 3 shows the Cherenkov angle distribution for all photons after a cut on dTOP of |dTOP| o2 ns for forward photons and |dTOP| o2.5 ns for backward photons has been applied. The measured resolution of 10.4 mrad is somewhat worse than our present Monte Carlo simulation's prediction of 9.1 mrad. The background under the peak comes primarily from ambiguous solutions to the photon direction. The refractive index of the radiator is a function of wavelength. This produces dispersion in the Cherenkov angle, with the red photons corresponding to smaller angles compared to blue photons. However, the red photons have a larger group velocity and therefore arrive at the detector before the blue photons. By measuring dTOP/Lpath, where Lpath is the total path length of the photon in the detector, one can infer the color of the photon and apply a correction to the measured Cherenkov angle. For forward photons, Lpath is typically 275 cm, while for backward photons Lpath ranges from 900 cm to 1400 cm. In Fig. 4 we show the measured Cherenkov angle relative to the expected angle vs. dTOP/Lpath for both forward and backward photons. As the forward photons travel a shorter distance, the correction is less effective than for backward photons. After applying this correction we see
Please cite this article as: D.A. Roberts, et al., Nuclear Instruments & Methods in Physics Research A (2014), http://dx.doi.org/10.1016/j. nima.2014.04.069i
D.A. Roberts et al. / Nuclear Instruments and Methods in Physics Research A ∎ (∎∎∎∎) ∎∎∎–∎∎∎
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66
3
Fig. 4. Chromatic correction for (a) forward and (b) backward photons. The plots show the shift in θC vs. dTOP/Lpath. The correlation is used to correct θC. Fig. 2. Measured dTOP¼ TOPmeasured TOPexpected distribution for (a) forward traveling (direct) photons and (b) backward traveling (indirect) photons. The fit curve is a double-Gaussian.
conclude that the detector design is robust to high background rates.
4. Conclusion
Fig. 3. Measured Cherenkov angle distribution for forward and backward photons after a cut on dTOP has been applied.
We have successfully constructed and tested a Focusing DIRC prototype detector utilizing a quartz expansion volume with a focusing element. Using approximately 600,000 cosmic muon triggers at the SLAC CRT, we have been able to demonstrate that the design performs as expected. Using the fast timing of the detector electronics, we have also demonstrated the ability to perform a correction to the Cherenkov angle due to chromatic dispersion. We will be publishing more detailed results of our analysis.
Acknowledgments an improvement in the Cherenkov angle resolution of approximately 0.8 mrad, compared to the Monte Carlo expected improvement of 1 mrad. To test the detector's performance in the presence of background, we injected light of variable intensity during certain portions of data taking. With a background photon rate of 144 MHz/bar box, we still see a clear Cherenkov photon peak. The peak resolution remains virtually unchanged but does sit on a larger background; the Signal/Noise ratio goes from 1.86 under nominal running conditions to 1.58 with high injected background photon rates. This background rate corresponds to roughly two times what the Belle-II TOP counter would expect. We
The authors would like to thank M. McCulloch for help in construction, final assembly of the optics and prototype, M. Benettoni for help in the design of the optics enclosure and mechanical support, M. Zago for the mechanical design of the Fbox and the Padova University mechanical workshop personnel for Fbox construction and assembly. We also thank M. Mongelli and V. Valentini of Bari for Fbox mechanical support structure design and M. Franco for helping to construct it, and M. Andrew of the University of Hawaii for electronics board design and support electronics. This work was supported in part by the U.S. Department of Energy, Contract DEAC02-76SF00515. Q2
Please cite this article as: D.A. Roberts, et al., Nuclear Instruments & Methods in Physics Research A (2014), http://dx.doi.org/10.1016/j. nima.2014.04.069i
D.A. Roberts et al. / Nuclear Instruments and Methods in Physics Research A ∎ (∎∎∎∎) ∎∎∎–∎∎∎
4
1 2 3 4 5 6 7
References [1] [2] [3] [4]
I. Adam, et al., Nuclear Instruments and Methods A 538 (2005) 281. C. Field, et al., Nuclear Instruments and Methods A 553 (2005) 96. J. Benitez, et al., Nuclear Instruments and Methods A 595 (2008) 104. K. Nishimura, et al., Nuclear Instruments and Methods A 701 (2013) 115.
[5] [6] [7] [8] [9]
J. Va’vra, et al., Nuclear Instruments and Methods A 639 (2011) 282. J. Va’vra, et al., Nuclear Instruments and Methods A 718 (2013) 541. L.L. Ruckman, et al., Nuclear Instruments and Methods A 591 (2008) 534. J. Va’vra, SLAC-PUB 13873 (2012). M. Borsato, et al., Nuclear Instruments and Methods A 732 (2013) 333.
Please cite this article as: D.A. Roberts, et al., Nuclear Instruments & Methods in Physics Research A (2014), http://dx.doi.org/10.1016/j. nima.2014.04.069i
Q3
1 2 3 4 5 6 7 8 9 10 11 12 13 14 Q1 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66
Contents lists available at ScienceDirect
Nuclear Instruments and Methods in Physics Research A journal homepage: www.elsevier.com/locate/nima
Results from the FDIRC prototype D.A. Roberts a,n, N. Arnaud b, B. Dey c, M. Borsato b, D.W.G.S. Leith d, K. Nishimura d, B.N. Ratcliff d, G. Varner e, J. Va’vra d a
University of Maryland, College Park, MD 20742, USA Laboratoire de l’Accélérateur Linéaire, Centre Scientifique d’Orsay, F-91898 Orsay Cedex, France c University of California, Riverside, CA 92521, USA d SLAC, Stanford University, Palo Alto, CA 94309, USA e University of Hawaii, Honolulu, HI 96822, USA b
art ic l e i nf o
Keywords: Cherenkov detectors RICH Photodetectors B factory
a b s t r a c t We present results from a novel Cherenkov imaging detector called the Focusing DIRC (FDIRC). This detector was designed as a prototype of the particle identification system for the SuperB experiment, and comprises 1/12 of the SuperB barrel azimuthal coverage with partial electronics implementation. The prototype was tested in the SLAC Cosmic Ray Telescope (CRT) which provides 3-D muon tracking with an angular resolution of 1.5 mrad, track position resolution of 5–6 mm, start time resolution of 70 ps, and a muon low-energy cutoff of 2 GeV provided by an iron range stack. The quartz focusing photon camera couples to a full-size BaBar DIRC bar box and is read out by 12 Hamamatsu H8500 MaPMTs providing 768 pixels. We used IRS2 waveform digitizing electronics to read out the MaPMTs. We present several results from our on-going development activities that demonstrate that the new optics design works very well, including: (a) single photon Cherenkov angle resolutions with and without chromatic corrections, (b) S/N ratio between the Cherenkov peak and background, which consists primarily of ambiguities in possible photon paths to a given pixel, (c) dTOP ¼TOPmeasured– TOPexpected resolutions, and (d) performance of the detector in the presence of high-rate backgrounds. We also describe data analysis methods and point out limits of the present performance. & 2014 Elsevier B.V. All rights reserved.
1. Introduction The DIRC detector at the BaBar experiment provided excellent particle identification performance [1]. Based on this success, we have been pursuing an R&D program to develop a compact and fast detector for future particle identification systems. One such idea, a Focusing DIRC (FDIRC) [2], would be capable not only of measuring the (x,y) coordinate of each photon with an angular resolution similar to the BaBar DIRC, but would also measure each photon's time-of-propagation (TOP) along the fused silica bar with 100–200 ps single-photoelectron timing resolution (the BaBar DIRC had a timing resolution of s 1.6 ns). The precise timing would allow one to correct for the chromatic dispersion contribution to the Cherenkov angle resolution and thus improve the angle measurement substantially ( 1 mrad). The improved timing resolution would also provide better background rejection, an important feature for detectors operating in the highintensity environment of a Super Flavor Factory for example, and
n
Corresponding author. Tel.: þ 1 301 405 6067; fax: þ 1 301 314 9896. E-mail address: [email protected] (D.A. Roberts).
could also provide modest improvement to the particle separation in some regions of phase space through the particle time of flight. By using photon detectors with a small pixel size one can reduce the size of the expansion volume by up to a factor of 10 relative to the BaBar DIRC while maintaining similar spatial resolution. The smaller geometric size together with better timing resolution would improve background suppression by nearly two orders of magnitude. The first prototype of the FDIRC concept was constructed and operated in a test beam in 2005, 2006, and 2007 [3,4]. This prototype utilized a single DIRC quartz bar and a cylindrical mirror placed in a mineral oil expansion volume. It was the first RICH detector to successfully correct the chromatic error using timing. Here, we describe a FDIRC design developed for the SuperB experiment that was to be built in Fascati [5,6]. Fig. 1 shows a GEANT4 realized image of the FDIRC, indicating key components.
2. Description of the FDIRC prototype The particle identification system for the SuperB experiment, or any other high-intensity Super Flavor Factory experiment, must
http://dx.doi.org/10.1016/j.nima.2014.04.069 0168-9002/& 2014 Elsevier B.V. All rights reserved.
Please cite this article as: D.A. Roberts, et al., Nuclear Instruments & Methods in Physics Research A (2014), http://dx.doi.org/10.1016/j. nima.2014.04.069i
2
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66
D.A. Roberts et al. / Nuclear Instruments and Methods in Physics Research A ∎ (∎∎∎∎) ∎∎∎–∎∎∎
utilize 1.5 μs of its 12 μs analog storage buffer. The digitized waveforms were analyzed off-line. The entire assembly was placed in the SLAC Cosmic Ray Telescope (CRT) [8]. The telescope features include: two planes of scintillator hodoscopes that provide 3-D tracking of cosmic ray muons with an angular resolution of 1.5 mrad and track position resolution of 5–6 mm; a quartz start counter providing start time resolution better than 70 ps; and an instrumented iron range stack used to select muons with energy greater than 2 GeV. The scintillator hodoscopes had active areas of 51 cm 107 cm and provided an angular tracking acceptance of 7 171 7271 in two respective directions. The bar box was located within the CRT such that the cosmic muons were incident near the midpoint along the length of the quartz radiators. The results to follow are based upon over 600,000 cosmic ray triggers in the CRT.
3. Experimental results
Fig. 1. The FDIRC optics as simulated in GEANT4.
cope with much higher luminosity related background rates than in the previous generation's experiments such as BaBar while maintaining similar physics performance. The estimated increase in background rates is on the order of 100-times higher. The basic design strategy for dealing with these increased rates was to make the camera much smaller and much faster. The design decision was to use a new photon camera-imaging concept based on focusing optics. The focusing element and expansion volume responsible for imaging the Cherenkov photons on to the PMT cathode surfaces were machined from radiation hard pieces of fused silica. In addition to the reduced background sensitivity gained from the geometric reduction in size, the use of fused silica also reduces background sensitivity, especially to neutrons, compared to a water-filled expansion volume. The improved timing not only helps with background suppression, but can also be used to measure and correct for the chromatic dispersion, thus improving performance. By including a cylindrical mirror that focuses light on to the photodetector plane we can remove the pinholesize (i.e. bar thickness) component of the angular resolution in one dimension of the quartz bar. The prototype was constructed utilizing a spare bar box from the BaBar DIRC. This is a complete bar box with 12 quartz bar radiators. The individual quartz bars had nominal dimensions of 35 mm wide 17.25 mm high 4900 mm long. The new optical components are a new wedge and a large focusing block (FBLOCK), both manufactured from solid fused silica [6]. The dimensions of the FBLOCK were nominally 560 mm high 217 mm long 422 mm wide. The new wedge was coupled to the bar box window with 50–75 mm thick epoxy. The FBLOCK was then coupled to the new wedge with 1 mm thick RTV. The entire assembly was housed in a light-tight mechanical support structure that connects to the bar box. The focal plane of the FBLOCK was populated with 12 Hamamatsu H8500 multi-anode photomultiplier tubes (MaPMT). This only partially covers the focal plane of the FBLOCK; complete coverage would require 48 MaPMTs. The location of the 12 MaPMTs was chosen to optimize the coverage for the expected hit locations from cosmic ray muons. Each MaPMT provides 64 pixels, with each pixel being approximately 6 6 mm2. The MaPMTs were read out with IRS2 [7] waveform digitizing electronics, preceded by a 40 preamplifier. The IRS2 was operated at a sampling rate of 2.7 GHz, and was configured to
To reconstruct the Cherenkov angle in a given event, we used a Monte Carlo generated dictionary that maps MaPMT pixel to photon direction inside the quartz bar, kpixel , as well as the time of propagation of the photon inside the FBLOCK. The dictionary was generated by simulating single photons in the quartz, drawn from an isotropic initial angular distribution. There may be more than one photon path that will lead from the quartz bar to a given pixel, giving rise to multiple solutions of kpixel for a given MaPMT hit. The dictionary is therefore multi-valued. Many of these ambiguous solutions can be eliminated either via timing information or because they give rise to unphysical values of the Cherenkov angle. The cosine of the Cherenkov angle was then calculated as the dot product of the track direction and the photon direction: cos θC ¼ ktrack Ukpixel [9]. The expected time-of-propagation (TOP) of a photon can be calculated from the track's entry location in the quartz bar as measured by the CRT, the photon direction kpixel , and the time-ofpropagation inside the FBLOCK taken from the Monte Carlo dictionary. This was then compared to the measured TOP, dTOP ¼ TOP measured TOP expected . This variable, dTOP, provides the primary background rejection and ambiguity resolving power. Fig. 2 shows the measured dTOP for forward traveling (direct) and backward traveling (indirect) photons. The difference in dTOP resolution between forward and backward photons is primarily explained by chromatic dispersion and is well modeled by the Monte Carlo simulation. Fig. 3 shows the Cherenkov angle distribution for all photons after a cut on dTOP of |dTOP| o2 ns for forward photons and |dTOP| o2.5 ns for backward photons has been applied. The measured resolution of 10.4 mrad is somewhat worse than our present Monte Carlo simulation's prediction of 9.1 mrad. The background under the peak comes primarily from ambiguous solutions to the photon direction. The refractive index of the radiator is a function of wavelength. This produces dispersion in the Cherenkov angle, with the red photons corresponding to smaller angles compared to blue photons. However, the red photons have a larger group velocity and therefore arrive at the detector before the blue photons. By measuring dTOP/Lpath, where Lpath is the total path length of the photon in the detector, one can infer the color of the photon and apply a correction to the measured Cherenkov angle. For forward photons, Lpath is typically 275 cm, while for backward photons Lpath ranges from 900 cm to 1400 cm. In Fig. 4 we show the measured Cherenkov angle relative to the expected angle vs. dTOP/Lpath for both forward and backward photons. As the forward photons travel a shorter distance, the correction is less effective than for backward photons. After applying this correction we see
Please cite this article as: D.A. Roberts, et al., Nuclear Instruments & Methods in Physics Research A (2014), http://dx.doi.org/10.1016/j. nima.2014.04.069i
D.A. Roberts et al. / Nuclear Instruments and Methods in Physics Research A ∎ (∎∎∎∎) ∎∎∎–∎∎∎
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66
3
Fig. 4. Chromatic correction for (a) forward and (b) backward photons. The plots show the shift in θC vs. dTOP/Lpath. The correlation is used to correct θC. Fig. 2. Measured dTOP¼ TOPmeasured TOPexpected distribution for (a) forward traveling (direct) photons and (b) backward traveling (indirect) photons. The fit curve is a double-Gaussian.
conclude that the detector design is robust to high background rates.
4. Conclusion
Fig. 3. Measured Cherenkov angle distribution for forward and backward photons after a cut on dTOP has been applied.
We have successfully constructed and tested a Focusing DIRC prototype detector utilizing a quartz expansion volume with a focusing element. Using approximately 600,000 cosmic muon triggers at the SLAC CRT, we have been able to demonstrate that the design performs as expected. Using the fast timing of the detector electronics, we have also demonstrated the ability to perform a correction to the Cherenkov angle due to chromatic dispersion. We will be publishing more detailed results of our analysis.
Acknowledgments an improvement in the Cherenkov angle resolution of approximately 0.8 mrad, compared to the Monte Carlo expected improvement of 1 mrad. To test the detector's performance in the presence of background, we injected light of variable intensity during certain portions of data taking. With a background photon rate of 144 MHz/bar box, we still see a clear Cherenkov photon peak. The peak resolution remains virtually unchanged but does sit on a larger background; the Signal/Noise ratio goes from 1.86 under nominal running conditions to 1.58 with high injected background photon rates. This background rate corresponds to roughly two times what the Belle-II TOP counter would expect. We
The authors would like to thank M. McCulloch for help in construction, final assembly of the optics and prototype, M. Benettoni for help in the design of the optics enclosure and mechanical support, M. Zago for the mechanical design of the Fbox and the Padova University mechanical workshop personnel for Fbox construction and assembly. We also thank M. Mongelli and V. Valentini of Bari for Fbox mechanical support structure design and M. Franco for helping to construct it, and M. Andrew of the University of Hawaii for electronics board design and support electronics. This work was supported in part by the U.S. Department of Energy, Contract DEAC02-76SF00515. Q2
Please cite this article as: D.A. Roberts, et al., Nuclear Instruments & Methods in Physics Research A (2014), http://dx.doi.org/10.1016/j. nima.2014.04.069i
D.A. Roberts et al. / Nuclear Instruments and Methods in Physics Research A ∎ (∎∎∎∎) ∎∎∎–∎∎∎
4
1 2 3 4 5 6 7
References [1] [2] [3] [4]
I. Adam, et al., Nuclear Instruments and Methods A 538 (2005) 281. C. Field, et al., Nuclear Instruments and Methods A 553 (2005) 96. J. Benitez, et al., Nuclear Instruments and Methods A 595 (2008) 104. K. Nishimura, et al., Nuclear Instruments and Methods A 701 (2013) 115.
[5] [6] [7] [8] [9]
J. Va’vra, et al., Nuclear Instruments and Methods A 639 (2011) 282. J. Va’vra, et al., Nuclear Instruments and Methods A 718 (2013) 541. L.L. Ruckman, et al., Nuclear Instruments and Methods A 591 (2008) 534. J. Va’vra, SLAC-PUB 13873 (2012). M. Borsato, et al., Nuclear Instruments and Methods A 732 (2013) 333.
Please cite this article as: D.A. Roberts, et al., Nuclear Instruments & Methods in Physics Research A (2014), http://dx.doi.org/10.1016/j. nima.2014.04.069i
Q3