Japanese KamLAND Site


List of Members
OD Status Reports

Apartment Schedule

PMT's status

Outer Detector Installation




Pressure Test Results of BNC Splicing Technique

The design of the cable splicing method we chose maximizes three important cost-saving activities. First, a large part of the cable preparation can be completed at our lab here in TUNL. Second, the PMT cable preparation will proceed using TUNL workers in Sendai and will not require assistance from non-scientists. Third, the assembly method is quick and easy, requiring about 15 minutes of time for one person.

While the design is cost-effective, we have not compromised water sealant and protection. High pressure water tests (described below) show that the seal will not leak, does not allow water to flow through the connection, and will not short the signal or high-voltage to ground. In addition, this method is simple and easily repeated with a very high reliability factor. This is very important considering that the procedure must be repeated 250 times and must provide a reliable electrical connection.

1. Cable splice design

The cable splice technique is based on the use of standard high-quality BNC connectors and epoxy-lined shrink-tube. The cable itself is "water blocked" with an inert substance inserted between the ground shield braid and the polyethylene outer jacket. This material will impede water flow in the cable if there is a hole or cut in the outer jacket insulation.

The splice design is shown in the mechanical drawings.

In the drawings, you will see the BNC connectors are crimped onto the PMT wire ends (RG-58 for the HV cable and 174 for the signal cable), and onto the new cable ends. The BNC connectors are the 2-part crimp style which is more reliable for electrical connections compared to the one-part crimp and the solder type of connectors. By crimping the ends with proper tools, we do not rely on soldering which can be inconsistant and subject to failure as a result of mechanical or temperature stresses. In addition, RTV-112, a silicone based epoxy (used for aquariums), is placed into the BNC connector between the outer crimp jacket and the grounded cable shield. This provides an additional water block between the outside and the inner connector. If water was able to flow through the blocking material to the connector, this seal will prevent this water in the cable shield region from entering the connector or the central pin region. In addition, any water flowing on the outside of the cable will be blocked from flowing into the connector region.

For the signal connection, we need to add a 50 ohm resistance to ground for proper termination. The use of BNC barrels (straight through adaptor pieces) with internal termination was prohibitively expensive (at least $30 each). We will use modified BNC T pieces. The Ts are cut to imitate a BNC barrel or straight through connector (see drawing). The chip resistor is soldered to the open end of the T. Epoxy will cover the resistor for protection.

The BNC connectors are easily pluged into the modified T piece, and a piece of adhesive-lined, polyolefin shrink-tube covers the complete connection. This shrink-tube is similar to that used in the Kamiokande cable splicing technique (see overall drawing).

For the HV connection, we will attach plug connectors on one end of the cable ends and a jack connector on the other end. The two connectors will then plug into each other directly.

We will protect the old cable ends on the PMTs by covering them with shrink-tube. In addition, we will place a metal ring (an electrical stand-off will be used for availability and price) on the 174 cable to provide an increased radius for the outer shrink-tube to seal.

2. Testing

The connection technique was tested using high-pressure air and high-pressure water. The BNC plug ends were crimped with RTV-112 sealant, onto the ends of normal (not water blocked or treated) RG-58 cable. A few cuts were sliced into the outer-jacket of the cable in one area. A Swagelock T and fittings were placed in the region of the outer-jacket cuts. This provided an attachment for an air hose and a water hose.

A picture of the test set-up is shown.

Approximately 60 psig Argon was put into the cable. There were no leaks around the Swagelock fittings. The BNC ends did not leak argon. When the pressure was increased to about 120 psig, one BNC end did leak. The test was repeated for high-pressure water at 40 psig. The BNC ends did not leak over the 3 week duration of the test.

These pressures are far greater than expected in the KamLAND water tank. In addition, the cable in the test was normal cable, and we will be using water-blocked cable.

We do not expect water to leak into the connector region as a result of water migrating from a pinhole leak upstream in the cable. We are confident that the shrink-tube wrapping will prevent water from leaking directly into the conector region.

3. Assembly

The new cable will be cut into the correct lengths and BNC connectors will be crimped and sealed into place. At the other end, the BNC signal connector and the SHV high-voltage connector will be crimped on. The BNC Ts will be modified, resistors will be soldered on, and an epoxy covering will be put on.

In Sendai:
The PMT cable ends will be cut to one length. The shrink-tube protection will be put onto these PMT cable ends. The radius piece will be put onto the small-diameter 174 PMT signal cables. The BNC connector will be crimped onto the ends of the PMT signal and high-voltage wires.

In the mine:
The wrapped new cables will be plugged into the PMT wire ends directly for the high-voltage cables and the modified Ts will be connected for the signal cables. The outer shrink-tube will be heated. This process will take approximately 10 minutes per PMT. The process will require a heat-gun. The cables will be labeled with an identification number and position.

The advantages of doing the final attachment in the mine are many. First, the cables can be shipped directly to the mine by surface which is much less expensive than shipping the cables to Sendai by air. Second, we will be able to drop the cables from the top and attach the ends to the bottom PMTs on the floor after they are mounted. This will decrease interference of personnel with the cables strung accross the bottom area. Third, the cables can be tested and replaced in-situ. Fourth, we can maximize our use of the long high-voltage cables, directing PMTs with short high-voltage cables to be placed on the top, midlength cabled PMTs on the sides and long high-voltage cables on the bottom.

4. Approximate Cost breakdown

$2,000      250 Ts for signal cable splice
$1,000      500 BNC connectors for signal cable splice
$1,000      500 BNC connectors for high-voltage cable splice
$1,450      heat shrink tube for connection
$   500      heat shrink tube for cable protection
$   250      RTV sealant
$6,200      total approximate (over estimate) cost for cable splice technique

Back to minutes.
Back to R & D.
Last updated: August 1, 2000