大面积阻性板探测器的研制_英文_

第27卷第7期2003年7月

高能物理与核物理

HIGH ENERGY PHYSICS AND NUCLEAR PHYSICS

Vol. 27, No. 7Jul. , 2003

Research and Development of Large Area

Resistive Plate Chamber

Z HANG Jia -We n 1; 1) ZHAO Ha -i Quan 2 LI Ru -Bai 1

DU Zh-i Zhen 1 XIE Y-i Gang 1

1(In stitu te of High En ergy Ph ysic s, The Chi nese Academy of Sciences, Beijin g 100039, China) 2(Beijin g Yinp eng Development Company of Science an d Techn ol ogy, Bei jing 100043, China)

A bstract We developed a ne w kind of material, which is used to construct the prototype of a large area resistive plate chamber (RPC) . In thi s paper, the structure of RPC, the curves of efficiency, counting rate, dark cu rrent, multiple hit and signal amplitude versus hi gh voltage measured using cosmic ray are presented. It shows that the RPC efficiency is close to 98%, counting rate and dark current are superi or to currently operati ng detectors world wide. Its performances well satisfy the requirements of high energy physics experi ments.

Key w ords resis tive plate chamber (RPC) , efficiency, counti ng rate, dark current

1 Introduction

Resistive Plate Chamber (RPC) is a type of parallel plate gas detector. RPC has superior time and spatial properties, which are e xcellent to measure the track and time of charged particles, and it can be built easily to

large area with low c ost. The resistive plate chamber is a new type of detector first developed by R. Santonico (Ro -ma) in the early 80. s . It has been used in several large scale detectors, suc h as the BELLE detector at KE K -B, the BaBar at SLAC, the L3at CERN, the CMS and ATLAS at LHC currently being built at CE RN and the YBJ -ARGO at Yangbajing International Cosmic Ray Ob -servatory in Tibet of China, etc. And so a lot of exper-i ences have been accumulated. B ES Ódetec tor is a large detector to be operated at the upgraded Beijing Electron Positron Collider (BEPC Ò). We plan to adopt RPC as BES ÓL detector. Before mass production is started, we developed a ne w resistive plate material, and used it to construc t a large area RPC prototype. The measured re -sults show that its performances well satisfy the require -ments of the experiments.

Received 15January 20031) E -mail:zhangj w@mail. ihep. ac. cn

[1]

2 General picture of R PC

RPC is c omposed of two parallel high resistive plate electrodes with a gap in between for the working gas to pass through. W hen a particle passes through this gas room, an avalanche or a streamer signal is produced. The

strips are placed outside the gas volume surface, and in -duced signals on the strips are read out. The mult-i layer RPCs allow simultaneous readout. The struc ture of RPC is shown in Fig.

Fig. 1. T he sketch map of RPC structure.

Since the material has very high resistance rate (10) 108#c m ) and the RPCs are independent from

615

616高能物理与核物理(HEP&NP) 第27卷

each other, the signals from mult-i layer hits increase the track de tection efficiency. In case the counting rate is high, only the avalanche mode is used because the stre -amer signal is larger than the avalanche signal, and so the volta ge drop is not recovered on time, leading to a lower efficiency. The advantage of the streamer mode just lies in its large signal, which does not need to be a mplified, so it can much reduce the cost of the electronics.

trigger signal. The area of all the 3scintillator detectors are 150mm @900mm. The two scintillator de tectors be -low the RPC are separated by lead brick, which removes the soft cosmic ray. The cosmic ray that crosses the 3scintilla tor detec tors at the sa me time must also crosses the RPC. The area of PRC is 600mm @1000m m. The gas mixing is completed by MKS mass flo w control system, with the three gases flowing into MKS 1259B mass flow controllers by the appropriate ratios. The RPC works in streamer mode. The high voltage is provided by the C AE N SY127high voltage systems. The positive voltage is ap -plied on the anodes and the negative voltage on the cath -odes. This approach minimizes the potential to the ground on c onnecters, cables, and surfaces and it serves as a precaution against external discharges through and around insulators. A computer controls the system, therefore the tests are completely automatic. Electronic system is par -tially NIM plug -in unit, while CAMAC and personal com -puter control syste m are used for the data collection and the high voltage control. The personal computer operates on Linux operating syste m and the control interface and graphic display are based on Root software development platform and C ++language environment. The system se-t up is shown in Fig. 2.

3 C onstruction of large area RPC

The crucial part of RPC is its gas room, which requires high level of technological process. It demands high degree of uniformities on the resistive plate, includ -ing the resistance, thickness, gap and graphite layer sprayed, especially it de mands very high degree of smoothness of the surface on resistive plate. Our technical process has been tested repeatedly before suitable resistive plate material is made. The resistive plate with thickness of 2mm is pressed with dense a mine film on the surface to guarantee its smoothness. Measured by the instrument of smoothness, its surface well matches the level of smooth -ness of glass. The bulk resistivity of the Bakelite is 4@108#cm at room tempera ture. The plates are separated by the circular insulation spacers with thickness 2. 0mm, 10mm in diameter and spanning 100mm from each other, and sealed at their surroundings by the strip insulation spacers with thickness 2mm. The mixed working gas, at a certain ratio Ar B F134A B Iso -butane =30B 62B 8, passes through the gap in between the two resistive plates from gas input to output installed on the opposite ends of the cha mber. Elec trically insulated Mylar separates the high volta ge electrode and pick up strips. The outer surface of the Bakelite is coated with graphite, which has the resis -tivity of 50k 8P t . The surface resistance is adopted such that it will not shield the signal produced when a charged particle crosses the RPC, and at the same time the resis -tance is smaller compared to the resistive plate, which en -ables the high voltage distributed on the entire resistive plate surface.

Fig. 2. RPC test system.

5 Test result

5. 1 Detection e fficiency

In the testing, the threshold value of discriminator is

4 Test system

3at

第7期张家文等:大面积阻性板探测器的研制617

ciency plateau curve versus the high voltage with different threshold values. This efficiency does not include the dead band of detector edge, but includes the loss of eff-i

ciency due to dead band of spacers, which takes 1. 25%.From Fig. 3, we find that the RPC reaches the efficiency plateau as the high voltage reaches 6. 8kV, and the eff-i ciency is close to 98%for the high voltage at 7. 2kV. The length of the efficiency plateau comes up to 1. 0kV. If two gaps of RPCs are used, the efficiency is virtually 100%, which completely satisfies the requirement of BES Ó. Threshold values below 200mV do not make visible differences at efficiency plateau. As threshold value goes above 250m V, the efficiency plateau obviously moves

backward.

Fi g. 5. The dark curren t versus high voltage.

rate apparently has a plateau. For the threshold value of

150mV, the counting rate stays around 0. 08Hz P cm on the efficiency plateau. As the high voltage goes up to 8kV, the dark current and the single rate increase rapid -ly. Compared with the experimental results worldwide, we find this result has come up to the level of glass RPC and the RPC with the resistive plate surface covered by linseed oil. From this we can see the smoothness of the resistive plate surface developed by us. The technological process is simple to master, which makes it easy to guarantee the smoothness of the resistive plate surface. 5. 3 Multiple counting

Fig. 3. The efficiency curve versus high voltage.

The multiple counting is defined as the ratio of the events with simultaneous signals from two adjacent pick strips to the total event number. It depends mainly on the value of the high voltage, the fron-t end electronic thresh -old and the gas mixture ratio.

To measure the multiple c ounting, the signal from pick strips is discriminated through the discriminator and is AND -ed with the signal from adjacent strip. Then all the AND -ed signals are OR -ed and afterwards AND -ed with the trigger signal before sending it to the scaler to count. The multiple counting is the ratio of this counting

5. 2 Counting rate and dark current

Fig. 4is the c ounting rate curve versus the high volt -age under different threshold values. Fig. 5is the dark

current curve versus the high voltage. The counting rate is the sum of read out from 4pick strips of 50mm widths. On a RPC, the counting rate on efficiency plateau be -tween individual strips deviates up to 15%.Here only the avera ge results of 4strips are

given.

Fig. 4. The si ngle counting rate versus high voltage.

The illustration is the same as Fig. 3.

Fig. 6. The multiple count versus high voltage.

T he illustrati on is the same as Fi g. 3.

618高能物理与核物理(HEP&NP) 第27卷

to the trigger signal c ounting. Fig. 6is the multiple count -ing curve versus the high volta ge under different threshold values. From this figure, we find that if threshold value is set to 150m V, in the middle of the efficiency plateau, the multiple counting stays around 10%, while if the thresh -old value is set to 50mV, the multiple counting can be as high as 60%.

5. 4 Signal amplitude

The signal amplitude is measured using the C AEN C420ADC. The signals of three scintillators are coincid -ed and expanded as the ADCs trigger signal. The signal from strips is delayed by 50ns and then sent into ADC in -put. The amplitude distribution chart is measured from 616to 9. 0kV, with a step of 100V. Fig. 7shows the av -eraged value of the signal amplitude versus the high volta ge. From it, we find that for the high voltage

lower

than 8kV, the signal amplitude is a linear function of the high voltage. This is the streamer mode range, and the signal amplitude is approximately 150) 300mV. At 8)

8. 6kV, the signal a mplitude increases rapidly. This is the transition point from strea mer mode to spark mode. As the high voltage goes above 8. 6kV, the signal amplitude no longer changes. In this range, it is seen from the eff-i

ciency plateau that the detection efficiency drops substan -tially.

6 Conclusion

We have successfully constructed a large area RPC using the material developed by ourselves. The length of the efficiency plateau e xceeds 1. 0kV and the efficiency is close to 98%.So if the high voltage is fixed in the cen -tral efficiency plateau, the efficiency of the detector is at ideal status.

With the ne w technology developed by us to make the material, its surface smoothness is comparable to the glass. This reduces the noise signal. The RPC produced with this technology has the good property of glass RPC, without the shortcomings such as the heavy weight of the glass, fra gility and erosion. It is also without the short -comings of linseed oil RPC that is unstable. At the same time, this technology makes it easy to guarantee the qua-l

Fig. 7. The signal amplitude versus high voltage.

ity and speed in mass production.

Reference

1 Santonico R. Nucl. Ins tr. and Meth. , 1981, 187:377

大面积阻性板探测器的研制

张家文

1; 1)

赵海泉李如柏杜志珍谢一冈

1(中国科学院高能物理研究所北京 100039) 2(北京市银鹏科技开发公司北京 100043)

摘要介绍了使用自制的材料制作的大面积阻性板探测器模型的结构和利用宇宙射线测量的效率曲线、单计数率曲线、暗电流曲线、多重计数和信号幅度随高压的变化曲线. 结果表明, 该RPC 效率接近98%, 单计数率和暗电流都好于国外的同类探测器, 完全可以满足通常实验的要求. 关键词阻性板探测器(RPC) 效率计数率暗电流

2003-01-15收稿