Action Items and Notes from the Meeting at Teledyne Electronic Technologies (TET)
February 11, 2003
Black: notes from R. Johnson
Blue, in parentheses: Additions
from R. Careyette
Present: Nick Virmani, Robert Johnson, Tom Borden, Brian Caplen, Rick Careyette
1. Nick will edit slightly his general QA document and give it to Robert as a starting draft to apply to the LAT Tracker subsystem.
2. TET will provide to Tom and Robert information on vendors suitable for doing the ASIC wafer lapping and dicing that TET has experience with. The dicing vendor should also do the visual inspection (look for chipped edges or scratches and verify that the correct version was put into the waffle pack). (Carol to supply information on our approved supplier base, i.e. Semi-Dice, who have the capability and resource to perform inspections in accordance with MIL-STD-883 Test Method 2010.)
3. TET will also provide to Tom and Robert a list of SMT parts distributors that they have had good experience with. (Carol to supply the TET Approved Vendor Listing for both passive and active elements.)
4. Rick will give to Nick a copy of the TET QA plan for class-H devices. (Copy of 7700530 MIL-PRF-38534 QA Program Plan, done.)
5. Robert will send to Brian before/after plots of the flex circuit so that he can verify the changes.
6. Tom will call Young Electronics and/or other PWB houses to get the ball rolling on that procurement.
7. TET will provide their Humiseal procedure to Nick, who will take it to Fred Gross at GSFC to obtain approval. Nick believes that it should be fine as long as it is cured at an elevated temperature to drive out the water. (Brian)
8. TET will also provide information on the die encapsulant material (Ablebond). Nick needs to take data and/or a sample to GSFC to get their approval. The cure conditions are relevant, so TET needs to supply that (80C was mentioned at the meeting). (Brian)
9. Nick will put together a spec for the flex pitch adapter.
10. Nick will get MCM #76 x-rayed, or some other method to try to see why the left-hand controller chip went dead during die encapsulation.
11. Tom needs to order the screw-head cups (or are they even needed now?)
12. Robert will make sure that the MCM PWB part number gets etched onto the boards.
13. Tom will send two connector savers to Brian.
14.Brian will figure out how many
times the MCM gets taken off and put on the carrier during manufacture. (Brian, define in
Unique Requirements document)
15. TET will provide Nick a view of their internal soldering processes so that he can work on getting the exception to NASA policy (regarding certification) approved. (Manufacturing and Quality are trained and certified to internal Process Procedures. The solder acceptance criteria is in conformance with IPC-A-610 Acceptability of Electronic Assemblies.)
16. Tom will add the solder paste to the materials list. Probably he needs the info from TET.
17. TET will update their process/procedure description for the MCM within about 2 weeks. (Brian)
18. Robert will document the lessons learned from the not-so-good MCM prototyping experience and distribute it.
19. Robert will start putting together a document for the MCM test procedure at TET.
Other notes:
· All MCM parts will be drop shipped to TET from the manufacturer or distributor or dicing vendor. A copy of the PO will be provided to TET in each case, and TET will check the shipping papers against the PO and receive the parts into their tracking system. (Treat as CFM, no incoming inspections required, count and ID only.)
· The PO for all SMT parts should clearly specify that only lead-tin be used (no pure tin).
· The SMT parts must all be put into reels by the distributor or manufacturer.
· TET will plan to do die sheer tests on a sample of MCMs. (Teledyne will do a "process qualification" on a given number of assemblies prior to final electrical since, they will be destructive types of tests on the assembly, component mounting and wire bonding.)
· TET will prepare extra boards with flex attached to use for destructive pull testing during the manufacturing process. Therefore, extra PWB and flex need to be ordered for this. (use of scrap materials for inprocess wire bond machine set up and process monitoring.)
· TET recommended Cirtech (Anaheim) and Fineline (Corona) for PWB manufacture.
· Nick said that 100% dimensional inspection is needed on the PWB (after machining the radius). He also wants some cleaning after the machining operation. The PO needs to specify how the parts are bagged (no paper with sulfer).
· A call was made to Omnetics regarding their nano-connector. Nick wanted to know about the space heritage of the contacts, insulators, and shell assembly. Omnetics said that the contacts are standard and used in all of their connectors. The connectors are mil spec tested but are not QMIL (since they are nano, not micro-D). The test data are available.
· Nick said that we need to have the PWB house do a complete design rule check on the photoplot data. They also must do 100% test of the final product to the netlist (not golden board).
· The PWB vendor should serialize the boards. The part number should go on the board layout, so that it is etched.
· The flex circuits need to be tested by repetitive flexing for 100 cycles.
· Nick said that no nondestructive pull test shall be done on any flight part. Everybody was happy with that.
· Robert would like to specify visual inspection of the tantalum capacitors after SMT mount, to be sure that none is reversed in polarity. (Brian to include in Unique Requirements document)
· Robert also wants to specify in the production flow that after SMT mount and connector mount (with savers), but before die attach, the MCM gets plugged into the power supplies and the leakage current gets measured.
· Robert noted that with connector savers attached the MCM doesn’t fit very well into the metal TET production trays.
· The PWB materials need to go onto the materials list at some point.
· Tom would like to make a shipping box that doubles as the burn-in fixture (holding 9 MCMs).
· The number of MCMs to build needs to be raised to about 800 to account for yield (this is a guess, of course).
· In case of yield problems, the process flow needs to include the possibility of rework by replacement of bad dice after electrical testing. The replacement will be done by stacking (removing the bonded die is difficult and hazardous). This may be done only once for a given die location.
· The UCSC test system will store and back up the data from the electrical testing (onto CD?)
· Homogeneous lot control by serial number needs to be carried out for the MCM production.