Minutes of the Tracker Technical Meeting
March 28, 2001
Sandro said the he talked to Plyform about a meeting there on the 24th of April. He cautioned that hotel reservations need to be made soon.
Sandro said that he will be in Terni tomorrow working on setting up the readout system for testing mockups in a climatic chamber. He is looking for a facility for vibration testing of mockups. Ladder and tray construction are under way with some small delay. At the beginning of next week he will have the first 2 trays without ladders made by Plyform. He also worked on closeout drawings, which were sent as dxf files to Hytec. He is waiting for drawings from Hytec to clarify ideas on clips. He requested 10 pieces from SGL of CC 1501 coated with SiC and machined. He plans to use some of the pieces in the mockups. He will send 2 to Hytec.
Sandro said that he would like to be able to cable the tower before removing any corner clips. This would require a narrow clip. Gwelen estimates 20mm available. BJ will send the tower drawings with cables to Pisa.
Erik reported that he sent 6-ply facesheets to Pisa on Monday, along with a sample of 4-ply face sheet and some samples of Alcomp CC material. Steve went with Erik to EDM Products to correct an error on machining of the assembly tools, which results from a misinterpretation of the Hytec drawings. Hytec still is using the incorrect jig but is getting a new version made up. Hytec is assembling a closeout and will test the ability to do the parylene coating before facesheet application.
Sandro asked what is the price of the Alcomp material? Erik estimated $400 to $600 per frame for material, assuming the large amount of machining on the backside for the facesheet alignment feature (which may be bonded on, instead, if that is found to be more cost effective). The parylene coating will cost about $20/frame. His wag for machining was $1000 per frame. That is for small quantitities, and he expects a sharp reduction for flight quantities. Robert requested that these numbers be firmed up as soon as possible. (Note: Erik later said that the production estimates from last summer still appear to be valid. My records show $1,300 per tray for the finished closeout pieces. These still need to be firmed up. R.J.)
Sandro said that SGL quotes $900 for a machined closeout following the simplified drawings.
Ossie reported that John packaged the glue application machine and is ready to ship it. The tungsten has been machined but is undersized by 2 mils. There is nothing going on with regard to the thermal testing. Omar will be back at the end of the week to document the results already obtained, which indicate no problem with stress in the silicon if the Nusil silicone adhesive is used.
Gwelen reported that the Nusil primer has no effect on live detectors when applied to the backside. There was a slight effect on leakage current observed when it was applied to the front side. He will now put it into the long-term testing setup. He reported that Nusil says that the primer should increase the conductivity significantly. UCSC will test that claim. Sandro noted that we have until the end of 2002 before gluing detectors down in production for flight towers, giving plenty of time for the long-term testing.
It was noted that there are issues on how to apply the primer. One idea was to spin it on, like resist is spun onto a wafer. We would like to keep it back from the edges of wafers. Gwelen suggested that we spray it on through a mask.
Ossie asked Sandro about problems that they have with using 3M 2216 glue for edge bonding. Sandro says that the bond is sufficient if they wait 3 days before removing ladders from the jig. After only one day they find that the bond is not sufficient to prevent modifications to the alignment by 20 microns or so. For final production, Sandro believes that having lots of jigs can satisfactorily solve this issue.
Gwelen reported that UCSC ordered pitch adapter prototypes from Q-flex, which quoted about ¼ the price of the previous quote (Tycho). He still would like to get some prototypes as well from CERN.
Gwelen has been looking at literature for information on metalization needed for high-yield wire bonding. For Q-flex he quoted 6um Ni and .5 to 1um soft gold, which looks consistent with the majority of commercial experience. The most significant yield problems arise from small pads sinking into the polymer. Unfortunately, we cannot make the pads any larger. There are some possible exotic options such as Titanium underlayer. Gwelen also noted that delay of bonding after touchdown can help on bonding to a soft substrate. He will document his findings.
Action Items: