Minutes of the Tracker Technical Meeting
May 9, 2001
Robert reported that the pitch adapters arrived back from Qflex. They had trouble keeping the long thin cover-layer aligned. Robert believes that this can be solved by extending the cover-layer between sets of 64 traces going to front-end chips, to tie it together. Sergei will revise the layout. Alec is working on bonding one of the pitch adapters to a dummy MCM board.
Tom will ask for a 14 mm increase in the stay-clear above the Tracker for readout cable clearance.
BJ reported that two cable mechanical layouts are done: the shortest (C0) and longest (C4). C1 and C5 are mirror images as far as the outline is concerned (but the routing internally will differ). He will finish the remaining ones when he gets an okay from Dave Nelson. Tom sent the layout to a company in Minneapolis, which can do a 24 by 36 inch panel.
MCM: Tom sent the gerber files down to Cosmotronic, for a quote on an aramid board. Dave Nelson will check on the price for an FR4 board. Tom will recontact Young electronics about possibly doing this board in aramid.
BJ just finished a more complete drawing of the bias circuit. Robert said that he liked Sandro’s layout of concentric circles for the gluing pads, to keep the dot squeeze-out symmetric. Tom will ask Sandro for a copy of his design in order to incorporate this feature.
Tom reported that he put the level-4 spec into Cyberdoc.
Nick Virmani is going to write up some assembly sequences or procedures as a straw-man for Tom and Sandro to revise into our PDR documentation.
PDR: Tom suggested 4 hours of questions and 4 hours for questions and discussions. About 1 hour should be allocated for Hytec. Robert will try to put together a more complete agenda. Hartmut will contact a DOE type (e.g. Helmut Spieler) as an external reviewer. Tom will contact an aerospace type to act as a reviewer.
Erik reported that he and Steve started writing a PDR design document. This document will reference the drawings. They will bring a drawing book to the review. Robert agreed to put the current status of PDR documentation on the web. (See http://scipp.ucsc.edu/groups/glast/manager/TKR_PDR/pdr.html)
Erik reported that Hytec released the closeout drawings.
Honeycomb: Hexcel will only sell unexpanded core in very big chunks: 20 inches thick. Another vendor (Plascore) will sell by slice, according to their documentation. Steve will verify this. 1 lb/ft3 appears to be standard, but not the 3 lb/ft3 stuff. Pisa wants to go to the proper thickness for their next mockups. Can Plyform order it themselves? The Hexcel cost is $2600 for the big block of 3 lb/ft3 core. Robert suggested that this is low enough just to buy it now and maybe use it eventually to do all of GLAST.
EMI shield: Steve talked to vendors for samples. He reported that the cost goes up fast above 50um thickness. Robert said that 50um thickness should be fine. The shield can still work well even if it is much less than a skin depth at the amplifier bandwidth.
Hytec is doing strength tests today on the carbon-carbon material. The coupons are ready to go.
We discussed the issue of thermal temperature drop on the tower. Hytec predicts 12.5C with measured value of sidewall conductivity. However, the prediction is16.3C with theoretical values for the sidewall conductivity. Erik is worried that the theory and measurement don’t agree for this case, since in all other measurements made (with other materials or the orthogonal direction in this material) they had good agreement. (The theoretical conductivity is 186 W/mK vs 255 measured.) Erik is worried about a possible mistake in the measurements, which would put us out of Martin’s spec. The prediction includes interfaces as well as closeouts and sidewalls. The contribution from the flexures is not included but is expected to be negligible. Martin is expecting an 8 to 12 degree drop. If we used P30 carbon-carbon (for which the measurement and theory agreed) the drop would be about 11C. Hytec assumed 0.45 W per MCM. (Note added in editing: this works out to 290 microwatts per channel, which is probably high by a factor of 1.5.)
Panel assembly tooling: Hytec will finish the drawings today or tomorrow. The machine shop already has the closeout drawings. The shop appears to be leaving a machine open, so we can hope for a short lead time.
Tungsten: there is no new information (except that Sandro found that the supplier he had in mind cannot supply the alloy in the form of foils that we need). Laminating two thin pieces of the pure tungsten may be the best solution. Must we us conductive epoxy? It may be possible to get what we want by grinding thick foils. Ossie will check more vendors of alloy. Robert said that Pisa needs thick foils soon for mockup trays. Thin material took about 1.5 weeks lead time to get at SLAC.
It was thought that we should ground converter foils to the bias circuit ground plan with conductive adhesive. Robert reiterated that he doesn’t like the 200 micron requirement on bias circuits but would prefer 50 microns. Dave is still holding out on IPC rules that have questionable applicability to polyamide flex circuits.
Action Items: