Miniproposals

Miniproposal: Date Filed: Title: First Author: Session Leader: Steve Wolfe (shots 1-26)

Engineering Operator Run Comment

flux loop test -- plasma?

Session Leader Plans

Physics Operators Plans

Session Leader Summaries

Entered: Jul 7 2004 03:49:09:790PM

Author: To Be Determined

Run Summary for Friday, 5/13/94 MP 041 - Restart CMOD SL: wolfe PO: horne EO: bertolino/fairfax The first two hours were devoted to demonstrating whether the flux loop problem was inside or outside the machine. We did several tests using a crude (but effective) winding outside the machine, connected to the inputs for F22,F23, and F25 using the same cable and female connector as is used for the magnetics. There was no sign of any problems with these signals, implying that the problem is inside. Shot#2 was a TF-only test, for the purpose of getting pickups for SNIPE's BP loops. The remainder of the run was an attempt to determine if we could in fact run with this many missing flux loops, or if it would be necessary to break vacuum and repair at least some. Shot 7 was the first attempt for plasma, done by recalculating all the magnetics predictors, ignoring F22 (as well as F23,F25,F17). This shot fizzled. Shot 9 was the first plasma shot, gotten by lowering the pressure and making the Br_0 1 mTesla more negative; this was a 400 msec shot, but the OH2's started doing funny things at around 200msec. After this shot, it was found that BP18_BC had caught the flux loop disease! We then struggled for a number of shots, trying to get another good shot. Several tweaks of fields and gas were made, mostly to no avail. Shot #12 went and lasted 400 msec. This was also the shot on which we lost BP25_BC. We continued struggling until shot 21, after which I became convinced that the early RCUR feedback was useful and needed to be done right (we had delayed this because it was using the bad BP loops, and so was doing the wrong thing.) The first shot after we swapped the BPxx_BC loops for the BPxx_DE loops into the Hybrid (and updated the Hybrid CONFIG, and recalculated all the magnetics predictors, again) was shot 22; this shot was a good plasma lasting until 230 msec, with problems beginning at 200msec. All subsequent shots ran, except 23 on which the pressure was too high. We succeeded in getting one shot that made it through current flattop and died during rampdown (#25) by turining off CLEARIN and ZXL (ZXL had gain=10.0000, which had previously been found to cause problems). Shot 26 had CLEARIN and ZXL turned back on, with timing and gains adjusted, and lasted until 470msec; this one may have died from too high an elongation. The question of whether we could run without the four missing flux loops was rendered somewhat moot by the fact that we continued to lose magnetics during this run, and there seemed no end in sight. All of the bad signals are on B-bottom flange, most on two connectors. While I think, given another day and stable magnetics count, we could have established satisfactory operation with the present magnetics set, the continued erosion of available signals makes a repair attempt advisable. The machine will be brought up to helium on Monday morning. No futher progress on the substance of MP 041 was made today. We still need to demonstrate gap and x-point control, and matching to the antenna limiter shape and position, to satisfy the stated goal of having the operation ready to carry out the physics program of Phase II. Altogether, we made four trips into the cell, stopping cooling and waiting for O2 to rise each time, and still got in 26 shots in an eight hours. Scorecard: Tests 6 Duds 6 Fizzles 8 Plasmas 6 ------- -- total 26

Physics Operator Summaries

Entered: Jul 7 2004 04:36:51:210PM

Author: To Be Determined

Run 940513 Physics Operator Report Horne (Phys op) Wolfe (Session leader), MP 41 Another unusual run. Last night I installed a 1 turn "flux loop" of perhaps .3 m^2 taped here and there under the machine. It was connected to Bob's integrators via his official cable -- the idea was to test all his circuitry for f25,f22,f23 -- hoping to reveal a fault in his rack somewhere. This was connected for shots 1-3 (one of these was tf only). The signal level was too small to verify operation, so we threw together a 50 turn coil (area 7 m^2) and installed it in place of the one turn loop. This gave a decent signal; all the integrators, etc seemed to work. To hammer in the last nail we used a cliplead to ground the end of this cable to the machine (another cell access) , and again observed good behavior. We concluded from this that whatever's broken is in vacuum; entered the cell again to reconnect Bob's cable to his feedthrough. We then proceeded to test the idea that we could run the machine using only the last miserable remnant of Bob's functioning flux loops. It became clear as we struggled that there were also two more bad BP loops connected to the hybrid, and we decided to break once more to replace them with functional loops from the DE set. This was our 4'th cell access today, and as a result we got plasma on 4 of the last 5 shots, loosing one due unusually high fill pressure (no settings were changed, we just got more gas than usual.) While our success on those last shots implies that we could probably run the machine in this configuration, the rather constant rate of failure per shot -- perhaps 1/10 of a diagnostic per shot, averaged over 3-4 runs, is very discouraging. The current plan is to do a helium backfill on Monday, pull the b_bot port and try to repair what we can. (All the problems are associated with this port.) One outstanding result of the run -- the work on the coolling system over the break has significantly reduced the nitrogen buildup in the cell, to the extent that quick access to the cell is possible. This combined with the fizzle detector reducing the coil heating on fizzles, allowed us to get both 26 shots and 4 trips to the cell today. shot 1 fake loop test 2 tf only 3 stack only Break for cell access to install multiturn loop 4 multiturn fake loop, all pf coils 5 multiturn fake loop, stack only Break for cell access to ground Bob's cable 6 multiturn connector grounded, all coils (except ef1, still warm) Break for cell access to remove fake loop and reconnect Bob's cable. recalculated predictors on wires 1-8 (after turning f22 off in hybrid tree) Offsets for these wires for next shot: wire name offset 1 rcur 10000 2 zcur 10000 3 zcur 10000 4 zxl 0 5 rxl 0 6 clearin 0 7 br -.0035 8 rcur 10000 br p gain at 5. 7 first plasma attempt; Flash. 8 dud 9 plasma 10 dud 11 fizzle. 1.2 Plasma is breaking down late. For next shot, raise ef4 i from -1020 to -990 12 plasma! 13 dud 1.2 delay clearin,dud 14 dud 1.2 15 fizzle 1.2 ef4 to 1005 16 dud 1.0 gas 17 fizzle 1.2 br -.004 (up .0005) 18 1.2 21 ef4 neg 15 A to -1020 Break for cell access; switch 2 bad bp_bc coils for de coils. Change hybrid tree, recalculate everything again. 22 good shot -- 220 msec plasma 23 turned z_xl pid off. It was at +10, exactly. Deleted a P point on rxl -- result is that it now turns on at gain of .38 and change High gas Fizzle 24 no change, good shot 250 msed 25 clearin gone (turned P gain off), 700 msec plasma! 26 zxl gain=8, 350, clearin 5 420, 470 msec plasma

Session Leader Comments
May 13 1994 08:25:40:750AM		Steve Wolfe	Run Plan for 940513: SL: wolfe PO: horne EO: bertolino First shot will be a test of the magnetics electronics, with the B/bot#2 connector removed from the machine and connected to a large loop. We may, depending on what we find, do a BT only shot to get pickup data on the SNIPES loops. If errors are observed on the first test, implicating electronics outside the machine, we will try to isolate the problem, then shut down so repairs can be done by next week. If no errors are observed on the electronics test, implying the problem really is inside the machine, we will reconnect the B/bot#2 connector and attempt to run, using new predictors that ignore F22, as well as F23, F23,F25, and of course, F17. The program will be to complete MP041. If running is successful, further analysis will still be required to determine whether we actually are able to diagnose and control the x-point region well enough to carry out the physics program. If we cannot get reasonable plasmas, or if subsequent analysis indicates that we really need the bad loops, we will need to have a vacuum break next week.
May 13 1994 05:40:56:910PM		Steve Wolfe	Erratum: In several places in today's notes I refer to B/bot#2 connector. The actual connector in question is B/bot#1 .
May 13 1994 11:12:26:740AM	940513007	Steve Wolfe	Shot 940513007: Back together and try to run. Do a call_alg on all magnetics predictors Fizzle detector 0.1 : 1.1 Turn Br0 gain back on. P= 1.3e-5 Result: Fizzle. Breakdown at 6.6msec
May 13 1994 11:27:09:320AM	940513008	Steve Wolfe	Shot 940513008: Drop the pressure to 1.0e-5 and go again. The previous Br looked a tiny bit positive, but leave it for now. Result: Dud. Br still looks positive to me.
May 13 1994 11:59:04:080AM	940513009	Steve Wolfe	Shot 940513009: Change BR0 offset from -3.5e-5 to -4.5e-3 Result: Plasma, Late breakdown - 10msec just over 400msec BP18_BC may have just lost it. The ZXL is driving the OH2's crazy. It has gain of 10.0000
May 13 1994 12:07:37:370PM	940513010	Steve Wolfe	Shot 940513010: Turn off ZXL gain. Result: dud
May 13 1994 12:27:03:600PM	940513011	Steve Wolfe	Shot 940513011: Turn off ZXL gain. Raise P to 1.2e-5 Result: fizzle - bd at 10.5msec Br looks a little negative; so does Bz.
May 13 1994 12:49:24:540PM	940513012	Steve Wolfe	Shot 940513012: Change IC_EF4 by +30Amps Result: Plasma. Lasts till 400msec. OH2U still goes back negative starting at 200msec. BP25_BC went dead on this shot.
May 13 1994 12:58:57:560PM	940513013	Steve Wolfe	Shot 940513013: Delay CLEARIN gain from 200msec to 350msec Result: dud
May 13 1994 01:09:12:860PM	940513014	Steve Wolfe	Shot 940513014: No change Result: Dud.
May 13 1994 01:25:41:730PM	940513015	Steve Wolfe	Shot 940513015: Put EF4 back negative by 15A, to -1005 Result: fizzle - breakdown at 10msec
May 13 1994 01:37:20:630PM	940513016	Steve Wolfe	Shot 940513016: Drop the gas to 1.0e-5 Result: dud
May 13 1994 01:55:57:120PM	940513017	Steve Wolfe	Shot 940513017: Raise gas to 1.2e-5 again Put Br_0 offset 0.5mT more positive, to -4.0mT Result: fizzle, breakdown at 10msec.
May 13 1994 02:09:05:680PM	940513018	Steve Wolfe	Shot 940513018: No change Result: fizzle- breakdown 12msec
May 13 1994 02:24:53:670PM	940513019	Steve Wolfe	Shot 940513019 Lower gas to 1.1e-5 Result: fizzle - bd at 11msec
May 13 1994 02:53:05:090PM	940513020	Steve Wolfe	Shot 940513020 Turn off RCUR feedback until 40sec (it's using the bad Bp's) Result: dud
May 13 1994 03:41:01:990PM	940513021	Steve Wolfe	Shot 940513021 Set EF4 to -1020A (15A more negative) Result: fizzle - bd at 11msec
May 13 1994 03:52:25:470PM	940513022	Steve Wolfe	Shot 940513022 Swapped BP18_BC and BP25_BC for the DE loops Modified the Hybrid CONFIG tree Did CALL_ALG on all relevant predictors. Result: plasma dies at 230 msec same nonsense on OH2's at 200msec
May 13 1994 04:07:36:770PM	940513023	Steve Wolfe	Shot 940513023 Turn off ZXL gain Move RXL gain to 380msec (delete first step) Result: fizzle, late breakdown - 12.8 msec
May 13 1994 04:23:13:500PM	940513024	Steve Wolfe	Shot 940513024 No change Result: plasma - short again 250 msec same nonsense at 200msec
May 13 1994 04:37:06:090PM	940513025	Steve Wolfe	Shot 940513025 Get rid of CLEARIN gain Result: Good long shot, around 700msec died during current ramp-down
May 13 1994 05:37:04:120PM	940513026	Steve Wolfe	Shot 940513026 Turn ZXL on gain=8 at 350msec and CLEARIN at g=5 at 420 msec Result: Plasma - lost it at 470m, may have just gone to too high an elongation.

Physics Operator Comments
May 13 1994 09:02:34:620AM	940513001	Steve Wolfe	Shot 940513001: This will be a test of the magnetics electronics. The B/bot#2 connector is removed from the machine, and F22,F23,F25 pins are all connected to a loop of wire. Br_0 gains are turned to zero in PCS, leaving IC control only. Inverts are set at 0.02 sec. Fizzle detecot is on from 0.1 to 2.1, as a backup. Result: Signal is only about 20 MWb on the 3 loops, a lot smaller than we expected.
May 13 1994 09:11:02:310AM	940513002	Steve Wolfe	Shot 940513002: TF only test. Fizzle detector changed to run 1.1 to 2.1 Result: well, at least something is working.
May 13 1994 09:11:35:830AM	940513002	Steve Wolfe	Shot 940513002: TF only test. Fizzle detector changed to run 1.1 to 2.1 Result: well, at least something is working.
May 13 1994 09:21:20:710AM	940513003	Steve Wolfe	Shot 940513003: Stack only, see if we get better signal. Result: Hardly any signal on those 3. We'll go down an try hooking up a multi-turn coil, and see what might be going on.
May 13 1994 09:26:31:350AM	940513003	Steve Wolfe	Shot 940513003: Note that EF3 got about 1000amps induced. Also note the other Floops with no signals include F01, F19,F21,F23, F25,F26
May 13 1994 10:00:01:570AM	940513004	Steve Wolfe	Shot 940513004: Change to about 50 turns .14m2 form. All PF's no TF Result: about 0.3 Wb, no evidence of the baseline not coming back at the end of the shot on any of the 3 signals.
May 13 1994 10:20:35:910AM	940513005	Steve Wolfe	Shot 940513005: Stack only. EF3 blocked. Result: Still no sign of a problem.
May 13 1994 10:30:50:020AM	940513006	Steve Wolfe	Shot 940513006: Try grounding the connector to the machine. Do all PF's except EF1's (they aren't cold yet) Result: No sign of a problem. It's inside, damn it!

Engineering Operator Comments
Shot	Time	Type	Status	Comment
1	08:53:14:780AM	Plasma	Ok	no faults
2	09:07:35:210AM	Test	Ok	no fault tf only
3	09:15:46:350AM	Test	Ok	oh1, oh2u, oh2l only - no faults
4	09:53:53:520AM	Test	Ok	no faults -- no tf
5	10:02:14:350AM	Test	Ok	no faults -- oh's only
6	10:25:25:090AM	Test	Ok	no faults
7	11:06:27:010AM	Plasma	Ok	no faults
8	11:17:56:150AM	Plasma	Ok	no faults
9	11:35:22:530AM	Plasma	Ok	no faults
10	12:00:59:960PM	Plasma	Ok	no fault
11	12:16:21:080PM	Plasma	Ok	no faults
12	12:30:13:750PM	Plasma	Ok	no faults
13	12:51:17:080PM	Plasma	Ok	no faults
14	01:03:02:040PM	Plasma	Ok	Dud. All else OK
15	01:18:30:800PM	Plasma	Ok	Fizzle. All else OK. Flux loops still dying.
16	01:34:32:740PM	Plasma	Ok	Dud. All else OK. Lowering gas did not help.
17	01:49:28:040PM	Plasma	Ok	Another fizzle.
18	02:02:49:490PM	Plasma	Ok	Another Fizzle.
19	02:17:25:050PM	Plasma	Ok	TF crowbar entering inversion; appears to be due to adrupt phase-in of
20	02:47:27:900PM	Plasma	Ok	dud. All else ok.
21	03:02:31:010PM	Plasma	Ok	All OK. Opening cell to work on flux loops.
22	03:47:10:030PM	Plasma	Ok	hot-wiring the magnetics rack gave a 600 kA plasma and 600 kA disruption
23	04:00:12:060PM	Plasma	Ok	Fizzle; late breakdown but almost went.
24	04:15:14:490PM	Plasma	Ok	600 kA plasma; disruption induced OH2L commutation failure
25	04:30:59:370PM	Plasma	Ok	600 kA discharge; disrupted at 500 kA on ramp-down; no faults
26	04:49:43:110PM	Plasma	Ok	600 kA plasma; 600 kA disruption at .47 sec