Some Mechanical Drawings for Layer00

We are constructing a mock-up of the beam-pipe and the carbon fibre sleeve inside of SVX. In parallel to this, we have started making mechanical designs for 8-fold and 12-fold geometeries, paying particular attention to how the detector will be installed. The life-size mock-up will allow us to test our designs.

We see four basic design strategies, driven mainly by the method of installation.

The full detector, consisting of modules, cables and hybids is constructed in two half shells. These are placed on the carbon fibre sleeve around the beampipe and slide into position, inside the SVX inner shell. The major difficulty with this solution is finding enough space to allow 4,6 or 8 hybirds to pass through the gap between the beampipe and the inner shell of SVX.

The detector is constructed in four quarters. Two are placed around the beampipe at each end, and slide into position. The advantage is that the hybrids do not need to pass through the gap. The disadvantage comes in the difficulty of uniting and aligning the two ends of the detector.

Another solution is to constuct the detector in two halfshells, but to place the hybrids on one end only. The advantage is that the hybrids do not need to pass through the gap. The disadvantage comes from the large lengths of cable required. This will firstly increase the noise, and secondly produce problems of space due to the large cable stacks.

A fourth solution consists of constructing the detector in two half shells, but without the hybirds. After insertion, the hybrids would be joined to the cables by means of a suitable connector.

The second and third solutions are unappealing for the reasons given above. We have not investigated the fourth solution because of doubts about what connectors could be effectively used. Therefore, most of our work concerns the first solution, and tries to arrange the hybirds such that they will pass through the gap, while maintaining the smallest possible cable lengths.

For the 8-fold geometry, a possible layour for the hybrids is as shown. The silicon is 256 strips wide and there are 3 segments in z. Each segment is readout by one master and one slave. The total number of hybrids is 96. The longest cable is 475mm. There is the possibility of mounting one master-slave pair on the narrow part of the beam pipe, thus shortening the cables by 80mm.
This is a different arrangement for the hybrids on the 8-fold geometry which may be required if 8 hybrids can not fit around the beam-pipe. In this design, only 4 hybrids encircle the pipe. All the hybrids which read out two 256-strip ladders are placed in a single row, half way between both ladders. This increases the longest cable to 545mm. There are some nice structural features about this design in terms of supporting the silicon and introducing cooling, but it would also require `dog-legs' to be placed in the cables, in order to make the connections.
A neater way of implementing the above ideas is shown here. Once again, only four hybrids encircle the beam-pipe, but every other pair of master and slave are offset and lined up with the ladder which they readout. The longest cable is 541mm.

We now turn to 12-fold geometries. Less hybrids (72) are required. However as the accompanying study shows this does not degrade the resolution: in fact it improves it considerably. A master and two slaves read out each ladder, one hybrid per z-segment. Each row of hybrids is aligned to a silicon ladder. Six hybrids fit around the beampipe. The largest cable length is 545mm. At the moment, this is our preferred solution. If one set of master-slave-slave can be placed on the narrow part of the beam-pipe, this will shorten the cable lengths by 110mm.
Alternatively, the hybrids can be arranged in a row, midway between the ladders. Structurely this has some benefits, but it makes the cable connections more difficult. There is now the freedom to read this out with 3 master-slaves, or 2 master-slave-slaves. The largest cable length is 475mm.

This is the mechanical arrangement for inserting a 12-fold detector. Six detectors fit around the pipe and inside the SVX sleeve. Note that this assumes the hybrids are 1.5mm thick. This number needs clarification. If the hybrids are much thicker, it may not be possible to insert them in this manner.
After insertion, the hybrids can be opened out and placed on a support structure as shown.

Noise

One worry we have concerning these designs is for the total noise. If the silicon plaquettes have an intermediary strip and are 15cm long, this gives 1.4pF/cm by 15cm =21pF. Add to this a cable of 54.5cm with a capacitance of .33pF/cm (54.5*.33=18pF) and you obtain a total capacitance of 39pF. From the performance of the Honeywell chip this gives a noise of 3000 enc after irradiation. This would give us a signal to noise of 7:1. This scenario is briefly discussed in our resolution studies and shows severe problems for the efficiency of an 8-fold design.

For the 8-fold geometry, a possible layour for the hybrids is as shown. The silicon is 256 strips wide and there are 3 segments in z. Each segment is readout by one master and one slave. The total number of hybrids is 96. The longest cable is 475mm. There is the possibility of mounting one master-slave pair on the narrow part of the beam pipe, thus shortening the cables by 80mm.
This is a different arrangement for the hybrids on the 8-fold geometry which may be required if 8 hybrids can not fit around the beam-pipe. In this design, only 4 hybrids encircle the pipe. All the hybrids which read out two 256-strip ladders are placed in a single row, half way between both ladders. This increases the longest cable to 545mm. There are some nice structural features about this design in terms of supporting the silicon and introducing cooling, but it would also require `dog-legs' to be placed in the cables, in order to make the connections.
A neater way of implementing the above ideas is shown here. Once again, only four hybrids encircle the beam-pipe, but every other pair of master and slave are offset and lined up with the ladder which they readout. The longest cable is 541mm.
We now turn to 12-fold geometries. Less hybrids (72) are required. However as the accompanying study shows this does not degrade the resolution: in fact it improves it considerably. A master and two slaves read out each ladder, one hybrid per z-segment. Each row of hybrids is aligned to a silicon ladder. Six hybrids fit around the beampipe. The largest cable length is 545mm. At the moment, this is our preferred solution. If one set of master-slave-slave can be placed on the narrow part of the beam-pipe, this will shorten the cable lengths by 110mm.
Alternatively, the hybrids can be arranged in a row, midway between the ladders. Structurely this has some benefits, but it makes the cable connections more difficult. There is now the freedom to read this out with 3 master-slaves, or 2 master-slave-slaves. The largest cable length is 475mm.
This is the mechanical arrangement for inserting a 12-fold detector. Six detectors fit around the pipe and inside the SVX sleeve. Note that this assumes the hybrids are 1.5mm thick. This number needs clarification. If the hybrids are much thicker, it may not be possible to insert them in this manner.
After insertion, the hybrids can be opened out and placed on a support structure as shown.