| Miniproposals | ||||||||||
|
| Operators | |
| Session leader(s): | Unknown |
| Physics operator(s): | Steve Wolfe |
| Engineering operator(s): | Bill Cochran,Gary Dekow |
| Engineering Operator Run Comment |
| Correction Coil Tests, Physop: Wolfe, Eng: Cochran/Dekow |
| Session Leader Plans |
| Physics Operators Plans |
| Session Leader Summaries |
| Physics Operator Summaries |
| Entered: Jul 7 2004 04:37:17:147PM |
| Author: To Be Determined |
| Physop Summary for 1030401 Power Tests of C-coils PO: wolfe EO: Dekow The goals of these tests were to 1) Verify safe operation of each coil, including a) lack of case arc'ing b) effect of fields on pumps, solenoids, computers, etc. c) adequate mounting of the coils to the igloo d) adequate clamping of all cables, leads, etc. e) adequate restraints on nearby magnetic or magnetizable objects 2) Verify/calibrate tmx operation with hybrid control 3) Evaluate field structure at the plasma for possible coil connection configurations; compare to calculated values *Requirements: 1) Patch panel complete and connected 2) Coil case clamps in place 3) All leads securely fastened down 4) TurboPump shields in place and secured 5) If possible, TMX current control non-linearity should be eliminated; otherwise we will need to compensate for it in the programming. Planned shot sequence: 0) Before energizing coils, carry out careful pre-ops inspection A) Remove and/or secure all magnetic materials in proximity to igloo B) Secure all cable leads to coils and patch panel C) Check hybrid and data system operation with no-power test [1 shot] We will need to limit in-cell activities for the duration of the power tests. It may be advisable to use the access control system, although we will need to cheat it in order to have in-cell observers. All personnel in the cell should remove magnetizable objects (like credit cards) from their pockets. Stray fields at ~1m from the coils will be in the 100 Gauss range. I) For each of seven coils Btop, Dtop, Dbot, Gtop, Gbot, Jtop, Jbot Pulse coil at low (1kA) and high (3kA) current [14 shots] For each shot: A) Visually monitor coil and nearby systems (people watching in cell) B) Monitor fields with gaussmeters C) Verify proper operation of all PLC/Paragon systems in the control room D) Observe signals on magnetics diagnostics E) Look for pick-up on all available diagnostic signals II) Measure coil and magnetics response in several series configurations A) "n=1 dipole" configurations 3kA [ 4 shots] 1) +Btop -Gtop 2) +Dtop -Jtop 3) +Dbot -Jbot 4) +Dtop +Dbot -Jtop -Jbot (4 in series may be voltage-limited) B) "n=1 quadrupole" configuration: [ 1 shot] +Dtop -Dbot -Jtop +Jbot C) Helical configurations [ 2 shots] 1) +Btop +Dbot -Gtop -Jbot 2) +Btop -Dbot -Gtop +Jbot Total of 22 "good" shots required. Results: -------- All tests were completed successfully. No case arc'ing was observed. There were no problems observed with any pumps, solenoids or gate valves, or computers. There was a small interference with the IGP1 ion gauge from the Jbot coil, but it did not turn it off. There were no problems with the igloo mounting of the coils. The first 3000A test of Btop coil opened the door of JR's rack. Some cable motion was observed with some of the coils, and will need to be addressed by improved restraints. These are noted in the logbook (user=dekow), and/or in the paper logs kept by the in-cell observers. Fields at approximately 1m from the coils were monitored with a gaussmeter and found to be roughly 100Gauss at 2900A. All shots were run under Hybrid control, which seems to work once the offset in the current regulator is compensated for; this seems to amount to a 700A offset in the demand waveform. The initial power test was run with a program of 1000A, because Bill had made a change at the tmx that was felt might remove the offset, but this turned out not to be the case. Subsequent attempts at 1000A were programmed for 1700, and for 3000A at 3700. Joe modified the PLC upper bound to allow this to be set to 4000 so as not to clip the 3700A hybrid programming. Operation with two coils in series resulted in lower current despite not having maxed out the supply voltage; this seems to indicate that the gain in the current feedback loop is too low. With four coils in series we were only able to get up to 2300A with the voltage clamped at 300V; for these shots, the drawn waveform was increased from 3700 to 4000, to make sure the demand caused full voltage, but this was probably not necessary. Increasing the demand might have helped with two coils in series, where the voltage was only 180V, but this was not attempted. Magnetics signals were obtained for all configurations, and should be useable for reconstruction of the structure, given knowledge of the coil geometry. The magnitudes of the signals were generally consistent with expectations based on previous calculations. Several potentially useful Bp loop signals appear to be broken, including some that were apparently working last campaign. These include bp16_gh and bp16_jk. It was noted that we have been using the bp16_bc loop in the processed flux, which implies that all of the other bp16 loops have had problems in the past; it is possible that these "broken" loops will recover once we start running disruptions. The bp16_gh loop actually appeared to have a signal, but it seems to have the wrong sign and magnitude. Note that the sense of the "winding" in my routine /home/wolfe/idl/cmod/error_fields/rui_coil is apparently opposite to the convention used for the actual coils, i.e. that positive current should correspond to a radially outward field. I'll need to confirm this and change it. All of the labeling in the cell appears to be correct and consistent with the above convention. Problems: --------- There was a slight delay in getting started while Josh fixed a problem with the announce routine (Pat's voice). This turned out to be caused by a permissions issue on the audio device, which actually was symptomatic of a more troublesome issue which is still being addressed. Apparently, when someone exits a session by lock file around that prevents subsequent logins, even by root, from obtaining ownership of a lot of /dev/* devices, with bad results. The first (no-power) shot was actually recorded as 1030325012 instead of 1030401001, due to a problem with determining current_shot('cmod'). The second shot was set up (by Josh) to be 1030401002, to avoid additional confusion. The magnetics integrators were turned off until shot#6, resulting in very small and puzzling responses to the first several coil tests. There was one cycle (shot 10) hang in INIT, apparently on an ECE action. This was quickly cleared by Josh. Shot #11 was supposed to be 1000A on Gtop coil, but was actually 3000A because I loaded the shot without applying the modified waveform in the all_v screen. That's the second time I've made that mistake in two weeks. Fortunately, nothing broke so it just ended up saving us a shot. On the positive side, the cycle seems to take about five to five and a half minutes, somewhat faster than last year, although we need to see how this changes with more scopes and more data being taken. Scorecard: Tests 24 Duds 0 Fizzles 0 Plasmas 0 ------ -- Shots 24 |
| Session Leader Comments |
| Physics Operator Comments | |||
| Apr 1 2003 07:53:53:840AM | Steve Wolfe | Getting set up for today's C-coil tests. The goals of the tests are : --------------------------- 1) Verify safe operation of each coil, including a) lack of case arc'ing b) effect of fields on pumps, solenoids, computers, etc. c) adequate mounting of the coils to the igloo d) adequate clamping of all cables, leads, etc. e) adequate restraints on nearby magnetic or magnetizable objects 2) Verify/calibrate tmx operation with hybrid control 3) Evaluate field structure at the plasma for possible coil connection configurations; compare to calculated values Requirements: ------------- 1) Patch panel complete and connected 2) Coil case clamps in place 3) All leads securely fastened down 4) TurboPump shields in place and secured 5) If possible, TMX current control non-linearity should be eliminated; otherwise we will need to compensate for it in the programming. The proposed shot sequence would be ------------------------------------ 0) Before energizing coils, carry out careful pre-ops inspection A) Remove and/or secure all magnetic materials in proximity to igloo B) Secure all cable leads to coils and patch panel C) Check hybrid and data system operation with no-power test [1 shot] I) For each of seven coils Btop, Dtop, Dbot, Gtop, Gbot, Jtop, Jbot Pulse coil at low (1kA) and high (3kA) current [14 shots] For each shot: A) Visually monitor coil and nearby systems (people watching in cell) B) Monitor fields with gaussmeters C) Verify proper operation of all PLC/Paragon systems in the control room D) Observe signals on magnetics diagnostics E) Look for pick-up on all available diagnostic signals II) Measure coil and magnetics response in several series configurations A) "n=1 dipole" configurations 3kA [ 4 shots] 1) +Btop -Gtop 2) +Dtop -Jtop 3) +Dbot -Jbot 4) +Dtop +Dbot -Jtop -Jbot (4 in series may be voltage-limited) B) "n=1 quadrupole" configuration: [ 1 shot] +Dtop -Dbot -Jtop +Jbot C) Helical configurations [ 2 shots] 1) +Btop +Dbot -Gtop -Jbot 2) +Btop -Dbot -Gtop +Jbot Total of 22 "good" shots will be required. | |
| Apr 1 2003 08:27:49:890AM | Steve Wolfe | Bring up PCS on cmodws12. I've got an acoil waveform programmed in seg2 from testing, but I'm going to go back to seg1 only for today. Just for fun, zero it in seg2 Turn off seg2 In seg1 all_v screen, program a waveform at 1800A, including 800A baseline offset I'll start the ramp at .1sec, up at .3, flat to 1.8 , down at 2.0 Load it. Try the coldstart from the Operations menu Still doesn't pidk up the full hloader log, but otherwise ok. All matrices respond. Tom's put in a line saying "Hit Last test, do a test shot #20 Looks ok. I'm getting 2.25V out on output 10. On shot 10324016 I got about 850A for a drawn waveform of 2V, so this ought to be ok. | |
| Apr 1 2003 08:32:35:403AM | Steve Wolfe | Change title in seg 1 from Startup to C-Coil test and reload | |
| Apr 1 2003 10:24:39:957AM | Steve Wolfe | Bill has made an adjustment on the tmx regulator, so I've redrawn the Acoil_cur waveform to 1000A to be safe. Bill will also set the plc limit to 1.2Ka as a limit. | |
| Apr 1 2003 11:04:00:020AM | 1030401001 | Steve Wolfe | shot#1: No power test It seems to be doing 1030325012 Didn't update the big board. bp17-20bc have a spike at -2 sec; all others ok. This was also showing up last week. Bob is going to check on the integrator board. |
| Apr 1 2003 11:16:52:790AM | 1030401002 | Steve Wolfe | 1030401 shot#2: No power test; Jas fixed something to do withe shot number. Big board updated this time. Magentics signals ok this time. I'm getting a hybrid demand 1.1V on the camac (with the factor of 2). Next shot : try for 1000A on Btop coil. |
| Apr 1 2003 11:30:29:830AM | 1030401003 | Steve Wolfe | shot#3: Try for 1000A hybrid control Btop only. PLC limit set to 1.2kA just under 300A on thecoil. Hardly any visible signal on the magnetics. next shot: raise the program by 700A, and see what we get |
| Apr 1 2003 11:53:11:253AM | 1030401004 | Steve Wolfe | shot#4: raise the program to 1700A, and see what we get; Increase PLC to 3kA. Got just under 1000A, as expected, but I still don't see anything on the magnetics. Bob sees up and down on f03 and f07, nowhere else. There may be an enhancement of noise on the upper BP loops. next shot: program to 3700A |
| Apr 1 2003 12:05:00:797PM | 1030401005 | Steve Wolfe | shot#5: program to 3700A; plc limit to 4000 Opened the door on JR's rack! Got 2870A on the coil. Seeing increased noise and some ac rise and fall on several de and gh coils. Also, f03 and f07 now show and on/off spike of +/- 10mWb I suspect the magnetics is not integrating or isn't hooked up right |
| Apr 1 2003 12:28:30:603PM | 1030401006 | Steve Wolfe | shot#6: repeat with integrators turned on! Now we have signals, at the mT level. Also we see signal on LM_CD and LN_KA. The polarity seems to be correct. Note that the coil sense in the rui_coil ruotine is backward. Next shot: Go to Dtop only at 1kA |
| Apr 1 2003 12:48:22:417PM | 1030401007 | Steve Wolfe | shot#7: Go to Dtop only at 1kA (program 1700) OK, polarity changed on the de_11 as expected. Looks fine. No motion observed. next shot: 3000A on dtop |
| Apr 1 2003 12:56:23:657PM | 1030401008 | Steve Wolfe | shot#8: 3000A on dtop Looks good, bp11_de =-3.5mT, makes sense. Saw one cable move; needs to get tied down. next shot: move on to Dbot at 1kA |
| Apr 1 2003 01:10:28:697PM | 1030401009 | Steve Wolfe | shot#9: move on to Dbot at 1kA small (~imT) negative signal on bp16_de, as expected. next shot: 3000A on Dbot |
| Apr 1 2003 01:19:49:430PM | 1030401010 | Steve Wolfe | shot#10: 3000A on Dbot cycle hung on an ece task; jas cleared it. Bill reports motion of the gas manifold line. Bp16_de=-3.5mT Next shot: Gtop at 1000A |
| Apr 1 2003 01:33:12:660PM | 1030401011 | Steve Wolfe | shot#11: Gtop at 1000A It came out at 3000A? I see a 1000A waveform in pcs But it looks like I never applied the all_v screen Physop error, that saves us a shot. next shot: Gbot, really at 1000A this time. |
| Apr 1 2003 01:44:07:470PM | 1030401012 | Steve Wolfe | shot#12: Gbot at 1000A (double checked) Signals look very small everywhere. next shot: 3000A |
| Apr 1 2003 02:11:04:820PM | 1030401013 | Steve Wolfe | shot#13: Gbot at 3000A Signal is positive on bp16_gh, may have the coil in backward. Bill says the coil is wired correctly. bp14_gh and bp15_gh are also positive, according to earlier shots should be opposite sign of 16 next shot: Jtop at 1000A. Sort the sign out later. |
| Apr 1 2003 02:13:52:097PM | 1030401014 | Steve Wolfe | shot#14: Jtop at 1000A bp11_jk -1mT, looks ok. note 12 and 13 are + next shot: 3000A on Jtop |
| Apr 1 2003 02:21:29:867PM | 1030401015 | Steve Wolfe | shot#15: 3000A on Jtop Looks ok, bp11_jk=-3mT No motion. Next shot: JBOT at 1000A |
| Apr 1 2003 02:33:23:723PM | 1030401016 | Steve Wolfe | shot#16: JBOT at 1000A I'm seeing nothing on BP16_JK. I think there may be a magnetics problem. Next shot: move on to 3000A. |
| Apr 1 2003 02:46:02:160PM | 1030401017 | Steve Wolfe | shot#17: move on to 3000A on JBOT Still no signal on bp16jk. Looks dead, have to get Bob to look at it after the run. BP12, 13jk are negative. So is bp17jk Note that we're using bp16_bc for the processed data, which implies that the JK must have been flakey at some point in the past. Some effect on the ion gauge IGp1. Next shot: Btop(+) and Gtop(-) in series at 3000A |
| Apr 1 2003 02:59:19:683PM | 1030401018 | Steve Wolfe | shot#18: Btop(+) and Gtop(-) in series at 3000A Current is lower, 2650A, V=180 Signals look reasonable. Next shot: +Dtop -Jtop |
| Apr 1 2003 03:14:54:647PM | 1030401019 | Steve Wolfe | shot#19: +Dtop -Jtop; I'll leave the programming the same for the two in series cases, but I'll probably try draing 4000 for the 4 in series to compensate for the poor voltage regulation. Magnetics signals seem good. 2650A, 180V. Next shot: +Dbot -Jbot |
| Apr 1 2003 03:30:06:433PM | 1030401020 | Steve Wolfe | shot#20: +Dbot -Jbot Looks ok, still getting ~2650A, 180V. Next shot: 4 in series, +Dtop+Dbot -Jtop -Jbot |
| Apr 1 2003 03:48:13:157PM | 1030401021 | Steve Wolfe | shot#21: 4 in series, +Dtop+Dbot -Jtop -Jbot; raise program to 4000A to try to boost the current demand a bit. This is expected to rail the supply voltage. Looks good. Voltage is pinned at 300V and we got 2280A, so that's as good as we'll get. Demand was indeed up, at 4.1V Next shot: +Dtop -Dbot -Jtop +Jbot |
| Apr 1 2003 04:04:03:697PM | 1030401022 | Steve Wolfe | shot#22: 4 in series, +Dtop -Dbot -Jtop +Jbot Looks good. bp12de - bp12jk and the 13 coils have 10mT of difference signal (n=1 component of Bz) BP15jk looks like it may have an amplitude problem? Current is holding at 2300A @300V. Next shot: +Btop +Dbot -Gtop -Jbot (one helicity or the other, need to work it out.) |
| Apr 1 2003 04:18:59:470PM | 1030401023 | Steve Wolfe | shot#23: +Btop +Dbot -Gtop -Jbot (one helicity or the other, need to work it out.) Looks different, but I can't tell much from the raw signals. n=1 signals are in the 5mT range, which is good. Move on to the opposite helicity: +Btop -Dbot -Gtop +Jbot |
| Apr 1 2003 04:27:36:503PM | 1030401024 | Steve Wolfe | shot#24: Move on to the opposite helicity: +Btop -Dbot -Gtop +Jbot Looks good. Still getting 2300A at 300V, the right signals changed sign. It all seems to work, except that we're missing some magnetics signals. |
| Engineering Operator Comments | ||||
| Shot | Time | Type | Status | Comment |
| 2 | 11:09:12:657AM | Test | Ok | Cycle Test |
| 3 | 11:24:32:787AM | Test | Ok | B top Correction Coil Test |
| 4 | 11:34:32:370AM | Test | Ok | B Top Correction Coil Test |
| 5 | 11:55:05:110AM | Test | Ok | B top Correction Coil test |
| 6 | 12:16:05:217PM | Test | Ok | B Top Correction Coil Test |
| 7 | 12:43:07:723PM | Test | Ok | D Top Correction Coil |
| 8 | 12:51:07:913PM | Test | Ok | D Top Correction Coil |
| 9 | 01:05:25:470PM | Test | Ok | D Bottom Correction Coil |
| 10 | 01:12:56:177PM | Test | Ok | D Bottom Correction Coil Test |
| 11 | 01:26:35:513PM | Test | Ok | G Top Correction Coil Test |
| 12 | 01:39:04:403PM | Test | Ok | G Bottom Correction Coil Test |
| 13 | 01:47:18:000PM | Test | Ok | G Bottom Correction Coil Test |
| 14 | 02:09:22:497PM | Test | Ok | J Top Correction Coil |
| 15 | 02:16:46:733PM | Test | Ok | J Top Correction Coil Test |
| 16 | 02:27:47:623PM | Test | Ok | J Bottom Correction Coil Test |
| 17 | 02:35:45:410PM | Test | Ok | J Bottom Correction Coil Test |
| 18 | 02:53:56:443PM | Test | Ok | +B Top, -G Top |
| 19 | 03:10:22:450PM | Test | Ok | +Dtop, -Jtop Correction coil test |
| 20 | 03:25:59:487PM | Test | Ok | +Dbot, -Jbot correction coil test |
| 21 | 03:41:15:307PM | Test | Ok | Correction Coil Test |
| 22 | 03:55:07:723PM | Test | Ok | Correction Coil Test |
| 23 | 04:10:31:077PM | Test | Ok | Correction Coil Test |
| 24 | 04:22:22:400PM | Test | Ok | Correction Coil Test |