Minutes of feedthrough meeting of 23 May, 2000. ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ Present: GV, Pof, MGF, PB, RM, RL, MLenc *) Milestones: - The six bellows that arrived 11 May have been leak checked with the (old) HLT160 leak checker; all six passed (and the HLT160 is functioning properly too...). - The (new) HLT270 leak checker has been mounted on the cold-test station and verified that it works. - The two LowVoltage Vacuum Cables and five 1.2 meter test harnesses that arrived the week of 8 May have been tested and are all ok. Margret will contact Axon to proceed with the balance of the LowVoltage Vacuum Cable order. - The second BNC BL-2 pulser has arrived. - Electrical tests of the 100 FCI vacuum cables delivered on 4 May have been completed. A summary of the test results is appended to the bottom of these minutes. *) BNL has just received eight 7-row pincarriers. We should be expecting an identical shipment now. *) Margret will send a reply regarding the EMEC request to `borrow' 12 spare pigtails, following Michel's recommendation that they purchase 12 extra pigtails for the test. ORSAY may, however, divert to EMEC 12 pigtails originally intended for UVIC and then, when they arrive, send to UVIC the 12 pigtails ordered by EMEC, as long as the diversion does not jeopardize our production schedule. *) Ebco has nine pieces of blank material left over from the flange machining. PB will pick them up when he is in Vancouver for ultrasonic cleaning. *) The heater-resistor wet/dry/freeze cycling continues, so far with no failures. *) The final thermal tests on the cold-test station will be carried out this week with (hopefully) the production heater design, including thermal conduction pads in place of thermal grease. *) Mark L, just back from CERN, reported on conversations with Gonidec regarding flange heater power supplies, fuses, power supply / heater interconnects, and thermocouples. Gonidec indicated that BNL is working on all of these. Mark will contact BNL for more information. Agenda for next week: ^^^^^^^^^^^^^^^^^^^^^ - Milestones - Warm flange heater design - when (or how) will we be satisfied that the design is final? - Thermal test results - Production readiness; critical path items and tasks to be performed in the interim before production - Baseplane issue - AOB ================================================================== Test results for the batch of 100 FCI vacuum cables received at the University of Victoria on 4 May: Visual Inspection: ^^^^^^^^^^^^^^^^^^ 98 cables passed, 2 cables failed. One of the failures was due simply to a malformed socket head on one of the jackscrews, making it impossible to insert a ball driver. The jackscrew can be easily replaced once we have received a supply of replacement jackscrews and retainer clips. The other failure was due to a lack of gold plating near the edge of one of the springclips. This cable is not repairable. Conductivity and Intermittent Fault Tests: ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ 100 cables passed, 0 cables failed. Each cable was tested with the Cirris cable tester operating in `continuous' mode, with the striplines being gently moved during testing. There was no evidence found of any intermittent faults. Impedance Tests: ^^^^^^^^^^^^^^^^ 100 cables passed, 0 cables failed. The impedance of six traces of each cable (three traces on each stripline) is measured with a network analyzer. All traces measured were within 33 +/- 4 Ohms as specified. Trace Resistance Tests: ^^^^^^^^^^^^^^^^^^^^^^^ 100 cables passed, 0 cables failed. The resistance of each trace of each cable is measured with a Keithley micro-ohmeter. The resistances ranged from 0.86 Ohms to 1.052 Ohms. Five pair of cables were identified whose resistances all lie within a 40 milli-Ohm band. This performance exceeds our requirements for calibration cables. Cross Talk Tests: ^^^^^^^^^^^^^^^^^ 99(*) cables passed, 0 cables failed. The peak-to-peak crosstalk on each of the two adjacent neighbors was measured for all traces. After subtraction of the crosstalk contribution from the interconnect cables, the measured crosstalk ranges from approximately 0.1% to 0.375%. This is within acceptable limits, and shows no sign of enhanced cross talk due to faulty ground traces. (*) The cable with the faulty jackscrew could not be tested since it cannot be inserted into a pincarrier until we have a replacement jackscrew. Contact Resistance: ^^^^^^^^^^^^^^^^^^^ First test -- 100 cables passed, 0 cables failed. The resistance is measured from each spring clip segment (13 segments on each springclip) to the corresponding spring clip segment on the opposite end of the cable. Performing the measurement for each side of the cable makes for a total of 26 measurements for each cable. On the initial contact resistance tests, three cables showed four `opens', five cables showed three `opens', and the other 92 cables showed two or fewer `opens'. All of these results are considered acceptable. Second test -- After all other electrical tests were completed, the contact resistance for a subsample of 29 cables was remeasured. Sixteen of these had five or more `opens', with two cables having more than ten opens. Note that the cables had each been cycled through one pincarrier insertion by this time. The problem seems to be associated with the way the top of the springclip hooks around the socket end of the connector. On compression, many of the springclip segments get caught over the top of the plastic connector housing and fail to spring back, thus loosing their springiness and becoming much flatter. We surmise that the springclip flattening might be happening while the cable connectors are being removed from the pincarriers after the cross talk tests, during which time the connector is rocked gently in a sideways motion to assist extraction. The problem does appear to be correctable. After carefully `unlatching' by hand the end of each springclip segment from the top of the connector housing, subsequent test results were consistent with the initial test results. Before final insertion into the feedthrough, we must be careful to ensure that each springclip segment is in its correct `relaxed' position.