Spectrometer Assembly Meeting (April 10, 1997):

present: Marek, Jurek, George, Gerrit, Heinz, Marge

Component List for the Spectrometer

The numbers given below include both arms. Table 1 gives the list for the default version of the spectrometer. In terms of production is has two defects:

• Frame 4 contains 1 module with a single sensor (Module 3). This means we need to design, produce and test a specific M-substrate for only 2 modules.
• Frame 5 contains 2 modules (module 8) where sensor types are mixed within one module (type 3 and 4). Furthermore one M-substrate on this modules is only partially equipped with D-substrates.

We discussed various possiblities to overcome these problems. The proposed version simplifies the production by reducing the number of different module and M-substrates

• Frame 4: The single-sensor module and 1 Module 2 is replaced by two new modules, which are identical to Module 7 but have type 2 sensors instead of type 3. This removes the single-sensor M-substrate.
• Frame 5: Module 8 can be simplified by using type 3 wafers in place of the two type 4 wafers. The resulting module is identical to Module 6, thus removes one type of module and all partially equipped M-substrates.

The simplification comes on the expenses of 50 additional substrates. It increases the granularity in frame 5 near the beam pipe and the acceptance in frame 4. Consequently frame 4 needs to be longer in the proposed version.

Staggering of modules

Some modules (Module 4 and 5) need to be at different distances from the frame. This can either be done by adding shims on each module or by incorporating it on the frames directly. Currently we assume to have is incorporated in the frames, so we do not need and addition production step in the module production and have identical modules for Module 4 and 5.

Gluing station

Jurek presented a first version of a gluing station: It consists of a fixed frame for the M-substrates. Two sensors are precisely aligned with respect to the D-substrates (bond pads) on two vacuum jigs, which can be shifted (X-Y) and rotated. Then the glass spacer is glued to M-substrates and sensors. After the glue cured, the vacuum jigs are shifted and the remaining (1 or 2) sensor(s) are aligned and glued in the same way. We aim on an overall gluing precision of 20mm to get good alignment between the bond pads on sensors and D-substrates. This seems to be necessary for the (automatic) bonding of the sensors to the D-substrates.

Other points:

• Question to Bernie: Do we need different flex cables for right- and left-handed M-substrates?
• In frame 8 we seem to have two places, where a screw, which holds the M-substrate to the frame, goes through a sensor, which is clearly not possible. In the default version, this problem also exists in frame 4. No solution to this problem yet …
• In the corners of frame 7 and 8 the covers protection the M-substrates reach into the sensors or neighbouring modules …
• Marek was concerned about the place for spacers to the cover (which should be in between the D-substrates). We should find some place on the M-substrate which is not yet used from SMD components …
• Piotr pointed out that, the use of the by-pass capacitor between back-plane and pre-amp input excludes any future possibility to readout the back-plane signal, as it couples the two sides.

 Heinz Pernegger, April 15, 1997