Minutes of the Tracker Technical Meeting

December 13, 2000

Present: Pisa, SLAC, Hytec, UCSC

Tom reported that he updated the MS-Project schedule. In particular, he linked the electronics assembly to the tray assembly. Everything gets assembled in lots of two towers. He noted that the ladder assembly is still missing from the schedule (it is assumed to be part of the tray assembly). Pisa needs to review the production schedule, put in detail for the ladder assembly, and make sure the schedule details are consistent with their resources and plans.

Robert and Masa reported that in addition to the BTEM MCM being used at UCSC for testing, there are two other complete MCMs. Masa tested them and found one to work fine. The other had a short from digital power to ground that had never been debugged since initial assembly. Masa traced the problem to a damaged GTFE64 chip. The chip evidently was gouged by a wire-bonder wedge during assembly. UCSC will repair the hybrid. In January somebody will carry them to Pisa to use for testing prototype tray assemblies. Sandro asked BJ to send a mechanical drawing of the BTEM hybrid so that they can prepare fixtures.

Gwelen reported on progress with the MCM design. The mechanical layout is close to being finalized, and we hope to begin the electrical routing next week. Some of the remaining questions are

1. How many screw holes? It looks like we can remove two for some marginal savings in cost.

2. Gwelen is concerned about the bias bond area, so BJ is working on a drawing of it.

3. Gwelen would like to move the chips away from the board end, since that is allowed by the new smaller size of the chip. The impact on the routing in the pitch adapter needs to be looked at. The new input pads are 200 micron pitch.

Sandro said that the way the present pitch adapter is laid out the traces between bonding pads are about 90 microns wide with 100 micron spaces.

Sandro reported that their wire bonder is limited to about 1mm in step height. It was decided to make the MCM mounting hole positions different in regular closeouts versus thick-converter-tray closeouts, so that the step from MCM to detector is the same in all cases. Otherwise we would be pushing the 1 mm limit.

We reviewed BJ’s new drawing of the cable routing on the side of a tower. Erik brought up an issue with the corner posts. He would like the 3.5mm dimension to be increased to 4.2mm to give sufficient material around a metallic bushing that needs to be bonded into each post. This means that the MCM needs to shrink by 1.4mm in length. BJ will make a new drawing for next week.

Erik presented some drawings of the Hytec flexure concept for mounting the tracker to the grid while avoiding stress from CTE mismatch.

Tom reported on a telecon held with a GSFC structural dynamics expert to get his take on results found in the SLAC analysis. He thought that we had plenty of margin in the fundamental frequency (35 Hz is the requirement, and 40 Hz was his level of comfort, while we have around 48 Hz with the flexure mount concept).

Erik reported that Franz sent the Hytec simplified Tracker model to SLAC this morning.

Tom reported that the original vendor contacted to work on the MCM and bias circuit decided not to support our R&D, because their production is running behind schedule. He contacted another vendor and sent our drawings to them. TRW came in with a very expensive quote on MCM assembly (about 5 times what we budgeted). Gwelen received questions from Teledyne but no quote yet. He has identified some more vendors to consider.

Ossie reported on some measurements he made on samples of the tungsten delivered. CSM foils are limited to 12 inches with a thickness tolerance of 0.7 mil. Those foils had a bright finish and their thickness (4.8 to 5 mils) was consistent with their tolerance. Schwarzkopf (sp?) can go up to 18 inches but with 5.5 mil tolerance on the thick converters. Their material had a dull, very coarse finish (bumps), and the thickness was from 24 to 32 mils. Ossie described a scheme to make a full-size plane with interlocked smaller pieces, which would allow us to use the better vendor. He shipped foils to Pisa, while retaining a few thick ones for prototype work at SLAC.

Ossie reported that he has received the new environmental chamber and new adhesive dispensing machine. He estimated late January for getting the machine understood and in operation. Robert said that UCSC would like to use it at that time to test encapsulation of the MCM. Ossie said that the machine comes with an offer of 3 or 4 days of factory training for 2 people.

Ossie reported that John is playing with glue volume versus dot size in preparation for making a dot-pattern sample for the thermal testing. There was some discussion of the pattern to use under the wire-bonding edge. Gwelen said it should be fine to keep it back about 3mm from the edge. He would like to use a bead instead of drops, so that it will act as a dam to keep wire-bond encapsulant from wicking under the detectors. Ossie asked whether adhesive tape might be used for that. Gwelen replied that with the present scheme for aligning the ladders before they contact the tray, adhesive tape such as the 3M conductive transfer adhesive originally foreseen (many years ago) for mounting of detectors could, in fact, be seriously considered. Ossie said that he had given adhesive samples to Hartmut to take to UCSC for aging tests of the thermal conductivity.

Nick reported that Pisa has made more tests of the DAQ, using the two BTEM trays. They have studied the noise occupancy versus position on the MCM and find the noise to be higher on the two ends. The plots are on the web. The effect shows up at low thresholds. At the 170 mV (30 DAC counts) threshold used in the beam test, they find the noise occupancy to be much less than 1/100000. They estimate the noise to be on average about 1100 electrons.

Sandro reported that they have continued to make ladders from glass wafers and encountered problems with the Scotchweld adhesive. They have now switched to a new sample of the adhesive. They are also wire-bond testing blank, aluminized wafers obtained from Micron and from Ossie. They will proceed in this direction toward tests of wire-bond encapsulation. He also reported that they have calibrated their measuring equipment and concluded that the Micron samples are 30 microns smaller than nominal.

Sandro reported that they have received an attractive quote from Laben for production of 16 ladders, including tooling. He will put new drawings of their tooling design onto the web this week.

Hartmut said that he would send out a PDF file on plans for detector testing. He is planning to have a review of the detectors by the end of January. He said that a P.O. is needed at HPK for purchase of the mechanical samples. He will give the price to Tom.

Decisions:

1. Make the drill pattern in the closeout pieces for the thick-converter panels different from that in the thin-converter trays, in order to keep the MCM wire-bonding surface the same distance from the detector bonding pads in both cases.

2. Increase the width of the corner posts on the electronics edges of the trays by 0.7 mm each.

Issues:

1. Can IPC rules be relaxed on the pitch-adapter trace width?

2. Should some screws be removed from the MCM design?

3. Dam scheme for preventing encapsulant of the MCM-to-detector wire bonds from wicking under the detectors.

Action Items:

1. Pisa: review Tom’s schedule for tracker assembly. Check that the time allotments are consistent with Italian planning. Add detail on ladder assembly.

2. UCSC: repair the second spare BTEM MCM.

3. BJ: send the BTEM MCM mechanical drawing to Sandro.

4. BJ: update the tray and MCM drawings to account for the new corner-post width.

5. UCSC: thermal testing of the silicone adhesive conductivity.

6. Hartmut/Tom: work out a P.O. for purchase of mechanical samples from HPK.