Full scale detector

After the verification of the principle with the prototype detector, a full-scale detector (hereinafter referred to as “LiNA TPC”) was designed as shown in Fig. 4.13 using 3D CAD software. Its basic structure is the same as the prototype detector. As shown in Fig. 3.1, this TPC has three detection layers to divide the detection volume to a signal and a background region. The TPC consists of ten wireframes and eighteen pillars made with nonmagnetic material of aluminum. The drift field is supplied by the drift boards mounted inside the TPC. The drift board has slits for the X-ray injection and a hole for the neutron beam injection. The electric signals on the MWPC are fed through twisted-pair cables to the amplifier boards mounted on the outside of the TPC. The specification of the TPC and its wires are summarized in Table 4.2 and Table 4.3, respectively.

Figure 4.14 shows a diagram of the high voltage supply to each wire. Anode wires have a low pass filter to cut off noise from the HV module. When the anode wire multiplies electrons with its high voltage, the charge is fed to the amplifier through a capacitor of 1000 pF. The drift field cage consists of a resistor chain (10 MΩ×13) to make a gradient of the electric field.

Fig. 4.13 Designed model of LiNA TPC. It consists of ten frames and eighteen pillars made with aluminum. Drift field boards are mounted on the inside and amplifier boards are mounted on the outside of the TPC.

Outer dimension 270×270×1020 mm³ Fiducial dimension 210×270×960 mm³

Anode voltage 1600-1900 V

Drift voltage 1000-2100 V

Operation gas He:CO2 = 85:15 kPa Table 4.2 Specification of LiNA TPC.

Anode Field Cathode

Material Au coated W BeCu BeCu

Diameter ϕ30±3µm ϕ100±10µm ϕ100±10µm

Direction x x z

Pitch 12 mm 12 mm 6 mm

Number 80 wires×3 layers 80 wires×3 layers 35 wires×6 layers

Bundle (mid.) 2 2 1

Bundle (up/low) 8 40 5

Table 4.3 Specification of LiNA TPC wires. All wires by The Nilaco Corporation.

“Bundle” means bundling number of wires connected to readout amplifier for middle layer, and upper and lower layers.

Wire mounting

The wires were mounted on the frame at a laboratory at Kyushu University. Figure 4.15 shows a schematic view of wire mounting, and the left side of Fig.4.16 is a photo of mounting. The wire mounting process is as follows. The edge of the wire is picked up from the reel and held temporary at the other side using a magnet. The wire is strain back to the reel side and put on a wheel. A weight is hung on the wire between the wheel and the reel. The wire is soldered on the board from the hold magnet side then the reel side. The total number of wires is 725: 160 wires × 3 frames for anode and field, and 35 wires × 7 frames for the cathode.

4.2 Time Projection Chamber 57

Middle Uppe

r

Lower

Cathode Anode

Drift resistor chain

10 MΩ ×13 HV module

To Amp.

Capacitor 1000 pF Resistor

1 MΩ

+VAnode

+VDrift

−VDrift

+VAnode+VDrift

+VAnode

+VDrift

}

Cathode

Low pass filter 1000 pF GND

Drift direction

Fig. 4.14 High voltage supply to LiNA TPC. HV module supplies to each layer of the TPC. Anode wires have a low pass filter to cut off noise from the HV module.

Drift field cage consists of a resistor chain to make a gradient of the electric field.

The white on black arrows indicate the drift direction of electrons.

Wire reel Wheel

Hold magnet

Wire mount board

Wire mount support frame

Weight Wire

Fig. 4.15 Schematic view of wire mounting.

Wire tension measurement

The tension of the wires was measured with the different method used for the prototype detector ^*2. The right side of Fig. 4.16 shows a photo of wire tension measurement, and Fig. 4.17 shows a circuit diagram of that. The middle point of the wire was sandwiched between two strong neodymium magnets (130 mT). Square pulse waves, whose height was +10 V and duty cycle was 15%, were sent to the wire from a function generator. The other side of the wire was connected to an oscilloscope and a resistor in parallel. The resistor converts current to voltage and the oscilloscope observe the voltage. One searches the resonance frequency with the dial of the function generator. The over-range squares are

*2That method may damage wires and require a quiet room.

Fig. 4.16 Photographs of wire mounting (left) and wire tension measurement (right).

input pulse waves and sine waves between them are an echo of the wire vibration. Note that, the wire vibrates not only with the eigenfrequencyf but also the frequency divided by integer,f /2, f /3, f /4...^*3. The measured tension of the cathode wire is summarized in Fig. 4.18. They are distributed ±10 g around the loaded weight of 100 g.

On the other hand, the anode and field wires loosen their tension due to the frame distortion. The left side of Fig. 4.19 shows the field wire tension at the first trial. The anode and field frame was distorted by their wire tension, then field wires loosen below the required tension of 20 g. Suppose a bar is loaded uniformly and both ends are fixed, the distortion δ is given as the following equation:

δ = wx²

24EI(L−x)², (4.2)

where w is a load, x is a position along the bar, E is an elastic coefficient, L is a length of the bar, I = bh³/12 is a moment of inertia of area, b is a width, and h is height^*4 of the bar. For the cathode frame the maximum distortion at x = L/2 is calculated to be δ = 2.6 µm using w = 16.7 kg/m, E = 70 GPa, L = 210 mm, b = 2 mm, and h = 30 mm. For the anode and field frame, the maximum distortion is δ = 1.14 mm using L = 960 mm. The distortion of the cathode frame was negligible, but that of the anode was large enough to loosen wires below the required tension. To suppress wire loosening, the weight on the anode and field wire was changed from 100 g to 50 g (δ = 1.14 mm → 0.57 mm), and the frame was pressured from both sides to distort forδ before mounting wires.

With these modifications, the anode and field wire tension satisfied the required tension as shown in the right figure of Fig. 4.19.

*3Recommended frequency isf /3 to observe clear vibration.

*4Parallel to the force direction

4.2 Time Projection Chamber 59

Neodymium magnet Wire

Function Generator Oscilloscope

Probe Probe

Resistor 100 kΩ

Pulse wave

Input GND

Fig. 4.17 Circuit diagram of wire tension measurement (left) and the oscilloscope screen of the vibrating wire signal (right). A function generator sent pulse waves to the wire which is sandwiched between two neodymium magnet. The wire vibration was converted to voltage signal at the resistor, and it was observed by the oscilloscope.

Cathode wire channel

0 5 10 15 20 25 30

Tension [g]

0 20 40 60 80 100 120 140

Cathode wire tension

Cathode No.1 Cathode No.2 Cathode No.3 Cathode No.4 Cathode No.5 Cathode No.6 Drift Bottom

Required wire tension 20 g Cathode wire tension

Fig. 4.18 Results of tension measurement for cathode wires. They are distributed

±10 g around the loaded weight of 100 g and satisfy the requirement.

Anode wire channel

0 10 20 30 40 50 60 70

Tension [g]

0 20 40 60 80 100 120 140

Anode wire tension

Field No.1 Field No.2

Required wire tension 20 g Anode wire tension

Anode wire channel

0 10 20 30 40 50 60 70

Tension [g]

0 10 20 30 40 50 60 70 80 90 100

Anode wire tension

Anode No.1 Anode No.2 Anode No.3 Field No.1 Field No.2 Field No.3

Required wire tension 20 g Anode wire tension

Fig. 4.19 Results of tension measurement for anode and field wires. The field wire tension at first trial (left). The anode and field frame was distorted by their wire tension, then field wires loosen below the required tension of 20 g. The results of tension measurement for anode and field wires with the pre-distortion system (right) satisfy the requirement.

Fig. 4.20 Construction process of LiNA TPC.The frames and pillars were stacked for three layers. The inside of the TPC was covered with drift field boards. The top of the TPC was covered with a clear polycarbonate plate.

Construction

Figure 4.20 are the photos of the construction process. The frames and pillars were stacked for three layers. The inside of the TPC was covered with drift field boards. The top of the TPC was covered with a clear polycarbonate plate.

High voltage supply

As shown in Fig. 4.14, high voltage is supplied to the TPC for drifting and multiplication of the ionized electrons. The stable voltage supply is an important factor of the detector operation. However, the detector discharged its voltage at the first high voltage test, but the location of the discharge was unknown because the TPC was in the vacuum chamber.

Therefore, a movie was taken by a camera through an acrylic flange to find a discharged position as shown in the left figure of Fig. 4.21. The right figure of Fig. 4.21 shows a discharged moment and the red arrow indicates the position. The discharge occurred between the components on the wire mount board. Thus, it was entirely covered by clear silicone rubber (TSE3032 by Momentive Performance Materials Inc., see Fig. 4.22). It has a high resistivity of 21 kV/mm. We had a lot of trial and error to achieve the required high voltage using the rubber and polyimide sheet.

4.2 Time Projection Chamber 61

Fig. 4.21 Detector discharge observation. A single-lens reflex camera was fixed with a tripod to take a movie of the detector discharge (left). The photo of spark discharge indicated by a red arrow (right)

Fig. 4.22 Bottles of sealing material TSE3032 (left) and the wire mount circuit covered with the clear silicon rubber (right).