One large area trigger detector for cosmic ray telescope system

The large area trigger detector is a key instrument in cosmic ray telescope system. One large area detector, sensitive size 1.3 * 2 m2, is proposed in this paper based on plastic scintillator tiles and wavelength shift optical fibers. Thanks to the wavelength shift fibers coupling to the scintillator tiles, only one photomultiplier tube is used to output the signals for whole large area detector. So this detector is simple and economy. The signal uniformity of this detector is better than 96% over the whole surface including the edges or corners. The detection efficiency of the muon is higher than 95%, and the time resolution is better than 10 ns over the entire detector. These performances are sufficient for the trigger detector in most cosmic ray telescope system.


Introduction
Particle detectors are often required to cover large sensitive areas, such as the ATLAS Resistive Plate Chambers (RPCs) [1] and the LHAASO electromagnetic particle detector [2].Cosmic ray telescope is a basic facility to check the quality or study the performance of such large detectors.The plastic scintillator usually is used as the trigger detector in the telescope owing to its fast-rising time, high light yield, low cost, and large area.Certain telescopes use only one small-size trigger detector to scan large area detectors, which is time consuming owing to low cosmic ray flux.Other telescopes combine several scintillator tiles as the trigger detector to cover the entire candidate large area detectors, which requires complex logical operations and electronics, such as the CORARS system [3].
Wavelength shift (WLS) fiber was used to collect and transfer the flash light from the plastic scintillator in many particle physics experiments [4,5].Such as the LHAASO electromagnetic particle detector is built with plastic scintillator and wavelength shift (WLS) fiber, wherein 4 plastic scintillator tiles were combined to a 1 m 2 area detector with the WLS fiber embedded in the grooves of the scintillator.It exhibits good uniformity over the entire detector area, the time resolution is better than 2 ns, and the muon detection efficiency is approximately 98% [6].
This study devised a scheme for the construction of a LATDE (large area trigger detector) by using plastic scintillator and WLS fiber, similar to the LHAASO electromagnetic particle detector.This trigger detector aimed to build one cosmic ray telescope to facilitate quality control of large area RPC of ATLAS upgrade project [7].Here, the maximum area of certain RPCs was larger than 2 m 2 .Thus, LATDE should fulfill the following requirements: (1) Sensitive area dimension: 1.3 m × 2.0 m.
The design scheme and the construction of LATDEs will be introduced in section 2. The performance evaluation of these LATDEs is presented in section 3.

Design of large area trigger detector
The design schematic plot for the proposed LATDE is shown in figure 1.Eight plastic scintillator tiles, each 1.3 m length and 0.25 m width, will cover the whole required sensitive area.In order to reduce the light loss from the WLS fiber end and keep good uniformity over whole sensitive area, U-turn and equal-length WLS fibers are embedded in scintilltors with equal-spacing.All the fiber ends are coupled to one same PMT for signal readout.The plastic scintillator type is HND-S2 manufactured by Gaoneng Kedi company, China (www.gaonengkedi.com).The scintillator tiles are produced by casting Polystyrene, p-Terphenyl and bis-MSB components.The peak emission wavelength of this scintillator is approximately 395 ~425 nm with quick decay time 2.4 ns, and the light yield was 50% of Anthracene crystal, more details can be found in the website of the company.The light decay length is approximately 2 m, but the WLS fiber will help to improve the light attenuation.In order to archive enough detection efficiency and light yield, 2 cm thickness of the scintillator is choice.Ω-type grooves are milled on the scintillator surface for embedding the fiber along the longitudinal direction as shown in figure 1.The spacing of the grooves is one key parameter which will effect the detection uniformity, light collection and also the time resolution.According to the simulation results, the spacing was optimized to 4.2 cm.
The light absorption peak of the WLS optical fiber should match well with the emission peak of the plastic scintillator.The absorption peak of BCF91A WLS fiber, manufactured by Saint-Cobain, is 400 ~450 nm and matches with the emission peak of Kedi scintillator.The re-emission peak of the WLS fiber is approximately 494 nm with decay time 12 ns.Its light decay length is longer than 3.5 m, which will help to improve the transmission of the light of the scintilltor tiles.In order to reduce the light yield at the fiber end, the fiber was inserted into one groove and U-turned in another alternative-two groove, both ends of the fiber are coupled to the same PMT as shown in figure 1.
-2 -All the fibers are cut at the same length to reduce the position effect on the signal time shift.Both ends of the fiber were polished using a fly-cutter with a coating of diamond on its edge.After balancing the cost and the performance, the 1 mm diameter and single cladding BCF91A fiber was chosen for coupling with the scintillator even the double-cladding and thicker fiber with higher light yielding.
Total of 24 fibers (48 ends) from the scintillator tiles are bunched together and coupled to the photocathode of a Hamamatsu PMT (CR284).The quantum efficiency curve of the PMT is well matched with the WLS fiber emission spectrum with low noise rate and quick response time.The gain of the PMT will be set according to the signal amplitude and the noise rate, following the gain and high voltage relation.Each tile was packaged using Tyvek as a reflector to enhance the light collection efficiency.All the scintillator tiles, fibers, and PMT were enveloped in a light-tight stainless steel box.Figure 2 shows the inner architecture of one LATDE detector without the PMT before closing the cover.Two LATDEs: LATDE1 and LATDE2, have been constructed and test.

Detector performance test 3.1 Test scheme
A cosmic ray telescope system was set up to test the performance of these proposed LATDEs.The schematic of this test system is shown in the left panel of figure 3. Two small area trigger detectors, 100 cm 2 regular hexagon plastic scintillator, are stacked up on the LATDE detector with a gap.This hexagon trigger detectors stack can conveniently move over the whole LATDE surface and scan the cosmic ray hit points on LATDE.The time resolution of these hexagon trigger detectors is better than 2 ns.
The output signal of each hexagon trigger detector is split into two channels, one channel input to discriminator and another to the digitizer for waveform recording.When one cosmic ray muon pass through LATDE and both hexagon trigger detectors, the signals of trigger detectors will pass the discriminator and produce one coincidence signal in logic unit.This coincidence signal will -3 -trigger the digitizer (FADC) to record the waveform of LATDE and both trigger detector signal as shown in left plot of figure 3.
All the data acquisition electronics of this cosmic ray test system were integrated into one VME and NIM mixed crate (CAEN NV8020A), as shown in the right panel of figure 3. The NIM modules, linear fan in-fan out (CAEN N625), low threshold discriminator (CAEN N845), and logic unit (CAEN N405), were used to split, discriminate, and coincidentally operate the output signal of the hexagon trigger detectors.One switched-capacitor digitizer (flash analog-to-digital converter, FADC) VME module (CAEN V1743) is used to record the signal waveform of the LATDE and hexagon trigger detectors, with sampling frequency capacity of up to 3.2 GS/s via 12-bit analog-to-digital (ADC).Further, there were 512 storage cells per channel, implying a 160 ns maximum recorded time per event.-4 -the fast-rising time was evident.The average amplitude of the first 100 points of the waveform was counted as the baseline of this waveform and will be subtracted in future data analysis work.
The uniformity of the large area detector is an important performance.Owing to the low cosmic ray muon flux, only 9 positions on LATDE (shown in figure 1) were scanned using this cosmic ray telescope in this paper.These positions included each corner and edge of the LATDE and central position.Positions 1, 2, and 3 were at the far direction from the PMT in different scintillator tiles, and positions 7, 8, and 9 were near to the PMT side and positions 4,5,6 are at the middle of each tile.

Performance test
The amplitude of the LATDE output signal was measured according to the waveform peak value (after subtracting the baseline).The distribution of the waveform peak at the testing position 1 shown in the left panel of figure 5 for 10,000 trigger event waveform.This distribution can be fitted by the Landau function convolving with a Gaussian function, the most possibility value (MPV) is 42.50 mV for the signal amplitude at position 1 of LATDE1.The MPVs of signal amplitude for all 9 positions of LATDE1 are shown in the right panel of figure 5.The average amplitude of all the 9 test points was 44.3 mV with the Root Mean Square (RMS) of 1.6 mV.The uniformity (1-RMS/mean) of this LATDE is 96.4%.The MPV of each position on LATDE2 is plotted in the same figure, and the average amplitude of all 9 test points was 44.9 mV with the RMS of 1.3 mV, and the uniformity is 97.1%.The signal amplitude keeps good uniformity even at the corners or edges for both LATDEs.
The cosmic ray muon detection efficiency of LATDE can also be obtained from the distribution of the signal amplitude, as shown in the left plot of figure 5.If the threshold of the LATDE is set to 10 mV, the fired events number  fired is the sum number of this distribution above 10 mV.Consequently, the detection efficiency is the fired event number divided by the total events  =  fired / all .The detection efficiency of all the 9 positions on LATDE1 and LATDE2 are plotted in the right panel of figure 5 also together the amplitude.The average detection efficiencies for all the 9 positions of LATDE1 (LATDE2) is 95.8% ± 0.3% (96.1% ± 0.7%) at 10 mV threshold.Due to both hexagon -5 -trigger detector stack up on same side of LATDE with narrow gap, there are some random coincidence trigger and induce the detection efficiency decreasing.
The time of the signal is another important performance and also can be measured with this cosmic ray telescope system too.When the system triggered, the waveform of all the LATDE and hexagon trigger detectors is recorded simultaneously.The time difference ΔT between the LATDE and trigger detector can be measured according to the time difference between the waveform at one threshold point.The time shift of the trigger detector is approximate 2 ns at 10 mV threshold, and will provide the reference time for the time measurement of the LATDE.The time difference ΔT between the trigger detector and the LATDE was measured at position 1 with threshold 10 mV, and the result is shown in left panel of figure 6.The time of LATDE signal is 22.47 ns later than the trigger detector, because the light of LATDE was collected (excited and emission) by the WLS fiber and transferred to the PMT while the trigger detector PMT directly coupling to the scintillator.The time resolution of the LATDE at this position is better than 6 ns.The mean value of ΔT for each testing position is shown in the right panel of figure 6.The maximum time difference between all the positions is approximate 8 ns as the light traveled different length in WLS fiber.The ΔT was nearly equal for the positions with the same vertical distance owing to the same length fiber, such as the positions 1 and 3 at different corners of the LATDE.Further, the time measurement results of LATDE2 were similar.Thus, the time resolution of the entire LATDE was 10 ns considering the combination of the position effect and the time resolution at the fixed position.If the hit position of the cosmic ray can be measured, such as with the aid of the candidate RPC detector, the time resolution can reach 6.0 ns.However, if the threshold of time measurement is decreased, the time walking effect will be reduced and the time resolution can be improved further.
Both LATDEs have been ready and installed in the cosmic ray telescope system for RPC detectors testing in the laboratory.The photograph of this system is shown as the figure 7. One LATDE is installed on the top of the system and another on the bottom, which just full cover the RPC detector in the middle.The performance of the LATDE meets the requirement of this telescope system.

Summary and discussion
Two large area detectors (2.0 m × 1.3 m) have been constructed by using scintillator tiles and WLS fiber.The performance of these detectors was examined by using one area 100 cm 2 hexagon prober.The uniformity of the single muon signal is better than 96%, including the corners of the scintillator owing to the fiber-improved light collection.In addition, the time difference between scintillator tiles was well controlled owing to the use of same length of fiber to transfer the light.
These LATDE had only one readout channel, which considerably reduced the cost of the electronics and the logical operation while maintaining good uniformity over the entire detector.Thus, the performance of these LATDEs is sufficient for use as trigger detectors for large cosmic ray telescopes, such as the ATLAS RPC quality control system.

Figure 1 .
Figure 1.The schematic plot of the large area trigger detector.8 scintillator tiles (gray block) are linked by U-turn WLS fibers (green line) and coupled to one same PMT.The fibers are embedded in Ω-type grooves as shown in the zoom plot.The WLS fibers are plotted only in first two tiles, the green dash line indicates the fibers embedded in other tiles.The number indicates the positions for performance testing.

Figure 2 .
Figure 2. Photograph of the inside architecture in one LATDe before closing the light-tight cover.Fiber has been inserted into the scintillators packaged with Tyvek.All the fiber ends will couple to one PMT later.

Figure 3 .Figure 4 .
Figure 3. (Left)The schematic diagram of the cosmic ray telescope test system.(Right) The photograph of DAQ system of this test system.

Figure 5 .
Figure 5. (Left) The distribution of the signal amplitude for position 1, the blue line is the fitted result by the Landau function convolved with a Gaussian function.(Right) Uniformity of LATDEs: the left axis is the MPV and the right axis is the detection efficiency of each position.The error bar is short than all the dots' sizes.Both LATDEs are plotted together for comparison.

Figure 6 .
Figure 6.(Left) The ΔT distribution at position 1 on the LATDE1.(Right) The ΔT distribution of 9 points of LATED1, the y-axis is the vertical distance between the testing point and scintillator tile edge near the PMT.The error bar is smaller than the size of the dots.

Figure 7 .
Figure 7.The photograph of the cosmic ray telescope system build with two LATDEs with RPC detector on the central layer.