From wagner@fnal.gov Wed Nov 21 15:04:46 2001 Date: Tue, 20 Nov 2001 16:03:21 -0600 From: Bob Wagner To: mukherjee@fnal.gov, Morris Binkley , JC Yun , ambrose@fnal.gov, burkett@cdfsga.fnal.gov Subject: COT wire pull procedure - what actually happened. [ Part 1, Text/PLAIN (charset: ISO-8859-1 "Latin 1 (Western Europe)") ] [ 329 lines. ] [ Unable to print this part. ] [ The following text is in the "iso-8859-1" character set. ] [ Your display is set for the "US-ASCII" character set. ] [ Some characters may be displayed incorrectly. ] Procedure to Extract the broken sense wire. Background: ------------------- -The broken wire is SL5-161-S11. It is broken about at the pultrusion and much is certainly "coiled up" at the WEST endplate, where P11 and P12 are shorted to ground (3.2ohm and 4.2ohm on WEST, respectively). The wire extends down and touches SL6-185-P0 somewhere on WEST side of the chamber (77 ohms from EAST and 14 ohms from WEST). -We wish to extract the wire from the chamber, cut it cleanly at the West endboard and pot the stub with corona dope. Since the stringing engine is not yet operational, we will use pullies attached to the survey IBEAM (the cell just happens to be at a convenient location for this). Preparation: ------------------- -Tue, 11/13: Mount WEST I-beam and measure distance to beamline and endplate. --Distance from surface of I-beam to beamline center: 10 3/8" --Distance from B0 side of I-beam to Quad side of Extrusion: 38" -->Calculate nominal pully height on EAST (low side) and WEST (done on 11/14 by Aseet) -Wed, 11/14: Practice removing a scrap wire plane from the prototype. -->11/14 17:43: One practice run was successfully done on the stringing prototype. At least one more practice run will be done on 11/15 before the procedure (JHA) is finalized. -Wed, 11/14: Measure the "Before" leak rate. Run a P5 line for "after" leak checks. (Del and company). -->??? Was this done ??? -- I will check in the morning of 11/15. -Wed, 11/14: Drain the SUVA. (Rob and company) Note that our electronics will be off until the repair is finished. -->11/14 08:50: 3v off to ASDQs at switches, caution tag about SUVA. -->11/15 Check with Dean or Bruce early morning for removal of line on WEST. ***IMPORTANT OVERSIGHT*** The TOF preamp and high voltage power (about 900 watts combined) was left on during the operation. Some outer cylinder temperatures approached 90 degrees and the endplate was typically in the high seventies. NEXT TIME BEFORE THE SUVA IS DISCONNECTED, THE TOF MUST TURN OFF AND TAG OUT THEIR PREAMP AND HV POWER. THE SUVA IS PROVIDING MOST OF THE COOLING FOR THE TOF. -Wed, 11/14: Pull power cord on USE LV filter box (kills 5v) -->11/14: DONE. -Lock out high voltage (keys in cabinet). -->11/14 08:50: Crowbar box keys removed and locked in cabinet. Procedure: ----------------- -Verify: High voltage locked out and power cord pulled from USE LV filter box. ---> 11/14 DONE. -Remove HVDBs and ASDQ DBs from area. Use judgement as to how many are necessary. HV daisy chain must also be moved out of the way. It's not clear whether or not ASDQ angle brackets can stay (likely not): If time, do this on Wed 11/14. --Suggest start by removing ASDQ DBs 5-158/159, 160/161 and 162/163; HVDBs 5-159, 160, 161, 162, 163. --> 11/14 DONE. -->AFTER: In order to facilitate the leak check, we also removed the microcoax for 156/157 and the HVDBs for 164 and 165. 165 is giving funny afterpulsing on all channels (a solid afterpulse), while 164 is pristine. My guess is that this is a "funny" calibration pulse combined with a high-ish noise level on the "odd" cell. We did not look at the calibration pulse. Note that no noise hits were observed on 165 when TRE and TRO = 0 and DTH still at 225mv. **During any part of the procedure, we must not allow conductors to fall into the chamber region. Back-up work with tape, etc., when necessary.** -Crack open the HV motherboard on the WEST and ASDQ motherboard on the EAST -- With hot knife, score a "break" in the bead along the extrusion. This will prevent the crack from spreading in phi. --Use chisels and such, as per Ken's instructions, to remove epoxy around MB. -Mount the pulleys at the nominal calculated position on the IBEAM (see above). The pulleys and jacks are mounted to unistrut uprights via angles, so that their coarse position can be easily changed. The fine positioning will later be done with the jacks. -Double check that the pulley washers are oriented properly (ie, the pulley really does lock). To lock the pulley, always tighten the wing nut WITHOUT the washer. -Speaker phones should now be hooked-up to allow good communication between the EAST and WEST endplates. Have walkie-talkies there as a back-up. -->AFTER: The speaker phones did not work due to excessive amibient noise on the EAST (voice activation always activated). We ended up using the walkie-talkies. It would be nice to do better (likely mics and speakers which don't rely on voice activation). I believe it would be worth the investment. -Attempt to shine the laser pointer through the slot ends to better align the pullies. (If this is difficult, a further procedure will deal with the problem below). -->AFTER: Aseet found that it was easier to align the pulley by sighting down it, through the near slot, and to the far slot (with lots of back illumination). Something like sighting down a rifle barrel. The high intensity lamp with gooseneck is very convenient for both fore and back lighting. I. WEST 1. Attach a safety string to board. -->AFTER: Wociech just hooked up the pulley wire / strings instead, since these did not seem to impede his work on the cell. He said that this was the usual way it was done at IB4. 2.CA a G10 piece to the endplate, separated to the endboard edge by the thickness of a parallel. (Later carefully clean off the CA). This will later allow the correct "re-seating" of the wire plane even if the "nib" is bad. Tape it to prevent accidental knock-off. -->AFTER: We found no place on the endplate to place this G10 piece. Instead, Aseet went back to the old transit check technique. He leveled the transit (a Brunsen that Kourosh borrowed from Alignment) and noted the offset from a convenient "pin base" on the wire plane above (5-160). This technique was re-producible at the 2 mil level. After marking it's position on the Ibeam / u-channel base, the transit was removed for the wire pull and replaced to later check the seating. It was actually used to correct the WEST seating. 3.Crack off G10 piece on wedge side of endboard. 4."Angle out" wedge while "angling in" temporary wedge to a perpendicular position at cell center. II.EAST Repeat I.1 to I.4 for EAST. III.BOTH EAST and WEST 1.Large portions of the EAST and WEST slots are now clear, so one can see through to the other side. Shine the laser through the opposite slot and better align the pullies on both WEST and EAST. -->AFTER: As noted above, the laser alignment was replaced by aiming down the pulley, near slot and opposite (backlit) slot. 2.Crack off epoxy between ledge and endplate. 3.Remove safety strings one at a time and attach stringing hooks to pulley. Attach 2.5 lb weights to each pulley. Shine the laser from opposite end and again adjust the pulley position. Lock all pullies. -->AFTER: This step was not needed, since the wire, string and weights were already attached. IV.WEST 1.Unlock the WEST pulleys. 2.Remove the wedge and pull the ledge free by hand (usual way). Adjust pulleys so that the endboard is free and centered in the slot. Lock the pulleys. V.EAST: Repeat IV for EAST. One may have to make another centering iteration on each side. -Look inside for the wire. -Unlock the EAST pulleys. -Have Mylar funnel available on the WEST side to help prevent wires from rubbing the edge of the slot if needed. -Put fingers on WEST pulleys (for a break) and unlock the WEST side. Rotate pulleys with thumb and index finger and start moving the plane from the slot. GO SLOWLY. Soon after moving, lock the WEST pullies and re-center the endboard in the slot. People on the EAST will make sure that the EAST endboard passes freely through the slot. -After the endboard clears the slot by a few inches, lock the WEST pulleys. -Pull the broken wire out of the chamber. Have an ohmeter handy to track progress (eg, does the short in 6-185-P0 disappear). Save the wire and measure the length to confirm that all is out. -->AFTER: It was confirmed that the length of the extracted wire stretched the distance between the endboard and pultrusion to within a few mm. -Snip the wire epoxy and pot the stub with corona dope. -Lock the EAST pulleys and unlock the WEST pulleys. -Use a Mylar "funnel" to help guide the EAST endboard into the slot. -Using the thumb and index finger technique, release the EAST pulleys and move endboards back into the slot. People on the WEST can keep their endboard well centered by hand. Some back and forth may be necessary. (If one could not tell whether the remaining 1/2 wire has reasonable tension, pull the East endboard through the slot and examine the wire.) -->AFTER: One could see the remaining 1/2- wire on the EAST, and it was "in-plane" at 5 lbs tension. We therefore did not pull the EAST endboard past the slot. -Latch the WEST ledge and install the wedge. Move the endboard against the G10 piece + parallel. -->AFTER: We actually latched the EAST endboard first (no critical reason - thought there might be slightly less stress on the remaining 1/2-wire doing it this way). The endboard nib was seated and later checked with the transit. Remember, the endboard is always seated toward the ID!! -Pull and latch the EAST ledge by hand. Install the wedge and move the endboard against the G10 piece + parallel. -->AFTER: The WEST endboard was actually latched last, and it's seating was checked (and corrected) with the transit. -Check for shorts with an ohmeter, before gluing the endboards back to the endplate. -->AFTER: No shorts were seen in 5-161. -Remove the G10 seating pieces and clean off the CA. -->AFTER: Obviously, not necessary using the transit. -On the benchtop, modify the ASDQ motherboard for cell 161 to tie S11 to P11. -->AFTER: For ease of later access (in case we wished to "float" S11), we actually tied S11 to P12. -On the benchtop, modify the HV daughterboard for cell 161 to remove the 200kohm series resistor for P12. Tie P12 to P11. -->AFTER: P12 was tied to P11 using an insulated jumper that avoided the solder pad for S11. Therefore, we could later float S11 without worry. Also, the 300 ohm series resistor for S11 was removed, so that the trace and wire stub on the WEST endboard would be floating instead of at the normal sense wire high voltage. -Attach the HV motherboard and check for continuity with an ohmeter. -->AFTER: Fine except for the expected "no-connection" for 5-161-S11. -Attach the ASDQ motherboards and check for continuity with the capmeter. Remember to put the dummy daughterboard in the ASDQ MB. -->AFTER: As expected, the 5-161-S11 capacitance was very small, and the P12 capacitance was double normal (hot legs of S11 and P12 tied at the ASDQ MB). -Epoxy the motherboards using DP-110 and the mylar strips (between boards). -Let set for 1/2 hour, and start flowing P5 at 60SCFH. -After about 1 hour, check for leaks and patch as necessary. -->AFTER: On the HV MB side, there was a "hot spot" at the ID extrusion to motherboard joint, and later one "old" hot spot was found on the OD after a couple more HVDBs were removed. All butt joints on both the HV and ASD side were "painted" using DP110 and an acid brush (a technique found to greatly improve that joint at IB4.) -Repeat the "after" total leak rate measurement with nitrogen, then flow nitrogen at 60 SCFH. -->AFTER: Unfortunately a gas line attached to the output donut was left dangling and unoticed after being disconnected from the beampipe. This caused a 3 to 4 SCFH increase in the measured leak rate. -Install the SUVA and get running again. -->Note that both the COT and TOF electronics should be off when the SUVA is disconnected. -Install electronics, doing the usual capmeter checks as we go. -->AFTER: As usual the neighbors were all checked, as well as the newly plugged boards. The capmeter readings for 5-161 P11, S11 and P12 were understandedbly unusual. With the ribbon disconnected, P11 and P12 read about 2.7 nf (their four 560pf caps plus that for S11 on the EAST end). The S11 reading reflected the disconnected sense wire. -Check SL5 and SL6 at 5% HV (hopefully there will be no "battery effects"). -->AFTER: Without the baggie, many SL5 channels drew DC current at 5% (moisture, uncured epoxy?). After two hours of baggie flow, the SL5 currents were very low at 5%. -Use Spy or the readout to check readout connections (see above). -Button up, install the baggie, wait until oxygen level is less than 1000 ppm before running up high voltage in nitrogen. -->AFTER: After flowing for about 2 volume exchanges the exhaust oxygen reading just started to dip below 2100 ppm (where the meter pins). This reflects the fact that the chamber was open to air for quite a while. We started HV tests at about 2000 ppm. SL5 and SL6 looked fine, but SL8-S11 (likely quad A) has developed a breakdown. This investigation is ongoing. Equipment Reminder -------------------------------- -G10 pieces and parallel to "re-seat" endboards. -->AFTER: Not needed - used transit check instead. -Hot knife, chisels for removing motherboards. -Cork/tube attachment for vacuum. -Little weights (add to 2.5lb for pulling planes). -Corona Dope -DP110 (two new tubes). -Nitril gloves -harinets -mylar strips for gluing MBs -Wedges -angles (Aseet) -Plane puller (Kourosh, if time) -mylar funnels (both 5 and 10 mil available) -->5 mil "funnels" (actually just "plane guides" on half the cell) worked fine. -Kapton (small sheets for protected wires in slots) -->AFTER: not needed. -CA (superglue) -->AFTER: not needed, except to attach temporay ledges on practise endboards. -Red leak detectors (syringe tips and vacuum?) -String and wires for pulling wire planes -Magnifying glass and flashlights. -->AFTER: The high intensity gooseneck lamps were extremely useful. -capmeter and ohmeter -Q-tips and Kay-drys. -->AFTER: also alcohol. -Speaker phones and walkie-talkies -->Speaker phones not very useful -- we used walkie-talkies