Miniproposals

Miniproposal: Date Filed: Title: First Author: Session Leader: Steve Horne (shots 1-23)

Engineering Operator Run Comment

Get plasma current to rise;mp41

Session Leader Plans

Physics Operators Plans

Session Leader Summaries

Physics Operator Summaries

Entered: Jul 7 2004 04:36:51:003PM

Author: To Be Determined

Physics Operator Summary for 5/4/94 SL: Horne PO: Wolfe EO: Fairfax/Gwinn Today's run was more noteworthy for problems than for solutions. We started from the last shot on Tuesday. Overnight Bob had switched the integrators on the suspect loop F25 and the unused loop F24. We quickly observed that F25 did not return to zero after the shot. During the course of the day, the magnitude of the difference between the pre- and post-shot baselines varied from -50mWb to -150mWb. The difference between the F25 reading at commutation and either of the adjacent loops also varied by from +50mWb to +100mWb. Note that these discrepancies are of opposite sign. We also encountered a problem on EF1L supply, such that the current went too high and the voltage did not follow the demand trying to pull it down. This seemed to be the case through shot #7. On shot #8 the engineers tried to diagnose the problem by clip leading in a CAMAC module to look at levels inside the regulator. The supply behaved correctly on shot #8 and thereafter. The clip lead was not removed. An attempt to get OH1 to get closer to programmed current by increasing the P gain on IC_OH1 led to a hesitation in the OH1 current ramp after commutation. This seemed to be due to accumulated integral error in the voltage regulator circuit, and went away when we dropped the P gain back to 1.0 The most troublesome problem seems to be the F25 loop. Reconstructions replacing this signal with the average of F24A and F26 gave VERY DIFFERENT flux patterns and Br values from those using the loop; BR_0 varied by of order 5 mT. Furthermore, examination of an old shot (931020003) indicated that F25 was about 20 mT more positive than either adjacent loop then. Replacing the signal by the average of F24A and F26 on that shot gave noticeably, though not radically, different recontructions. Bob points out that there is finite current, of order 10kA, in the divertor hold down structure near the F25 loop, and that this structure is not up-down symmetric. Therefore a small difference between F25 and the near neighbors is not ruled out, and a satisfactory replacement scheme is problematic. On shot 13 we loaded a Br predictor replacing the F25 by the average of F24A and F26. On shot 16 we got a late breakdown using this predictor, and after some tweaks to BR_0:offset and to IC_EF4. An attempt on shot 20 to re-run shot #8, which broke down using the original predictor, resulted in no breakdown. It was noted that the F25 signal differed more from F24A,F26 on shot 20 than it had on shot 8. Replacing shot 16 and tweaking EF4 again resulted in a reasonable breakdown and 30kA, starting at 9msec, on shot 23. The MFLUX reconstruction based on replacing F25 showed a passable null early but no null at 9msec when it actually broke down. What a mess! After the run, Bob and I noted that on shot 22, which was the last shot that did not have a TF crowbar, F23 also had unpleasant symptoms, namely a 40mWB level at the end of the shot. Bob insists that F23 is not shorted to the vessel at all, and suspected integrator drift. A test shot showed no integrator drift, at the bit level over 4sec, on any flux loop. Nevertheless, Bob put the F25 loop back on its original integrator circuit, and swapped F23 with the unused F24. We'll watch it tomorrow. Good things (a short but non-null list): Despite some stoppages for the EF1L problem and lots of thinking time, we got in 23 shots. The LN2 pump was run as high as 30Hz with no sign of a N2 leak. Score card: Duds 15 Fizzles 8 Good 0 ----------- Total 23

Session Leader Comments

Physics Operator Comments
May 4 1994 09:36:36:040AM	940504001	Steve Wolfe	Shot # 940504001 ----------------- Start from last shot yesterday (reloaded it just in case) Only changes to the system should be the switched integrator on F25. Also, Joe will switch the bus voltage taps the on the OH2 bus voltage that was reversed yesterday. P=1.0e-5 Torr D2 Result: No breakdown
May 4 1994 09:48:02:430AM	940504002	Steve Wolfe	Shot # 940504002 Raise P to 1.3e-5 Result: Still no flash. Br may be too negative. F25 (on the new integrator) is not coming back to zero.
May 4 1994 10:16:26:730AM	940504003	Steve Wolfe	Shot # 940504003 Raise Br_0 offset from -.002 to -.001 Result: Got a flash. we seem to be 1mT different from yesterday. Breakdown at 10msec. Null looks small.
May 4 1994 10:39:16:300AM	940504004	Steve Wolfe	Shot # 940504004 Raise gain on IC_OH1 from 1.0 to 2.0 and eliminate two early points. OH1 current was not reaching target. Result: flash - breakdown at 11msec, hardly any current. No burnout. OH1 faulted late.
May 4 1994 11:00:51:470AM	940504005	Steve Wolfe	Shot # 940504005 Change IC_OH1 tfrom -16600 to -16100 (previous move may have been wrong way) Result: No breakdown. Null looks terrible.
May 4 1994 11:21:02:830AM	940504006	Steve Wolfe	Shot # 940504006 Change IC_OH1 back from -16100 to -16600 (previous move was wrong way) This should be the same as #4. Result: No breakdown???
May 4 1994 12:12:17:860PM	940504007	Steve Wolfe	Shot # 940504007 Put BR_0 offset back to -0.002 Result: Got the flash back. Breakdown at 7msec ON previous shot, noted that EF1L is overshooting it's programmed current. It looks like it has an oscillation, and is not following its demand (wrong sign of voltage).
May 4 1994 12:45:47:790PM	940504008	Steve Wolfe	Shot # 940504008 Re-do same thing with some instrumentation on the EF1L supply For the hell of it, P=1.0e-5 Result: Still got a flash breakdown at 5.65msec This time the EF1L worked. Null including F25 looks fair, but with F25=(F24A+F26)/2. it looks horrid!
May 4 1994 01:15:21:600PM	940504009	Steve Wolfe	Shot # 940504009 Change VOH1 to go positive right after commutation (was held negative until 4msec Result: NO breakdown
May 4 1994 01:24:13:760PM	940504010	Steve Wolfe	Shot # 940504010 Put P gain on ICOH1 back to 1.0 and change VOH1 back as it was. We suspect the problem is the integral gain. Result: No breakdown There is no null. The Bz evolution and OH1 evolution is better, but radial field locks awful.
May 4 1994 01:36:06:910PM	940504011	Steve Wolfe	Shot # 940504011 Change BR_0 offset to -3.e-3 Change P=1.2e-5 Result: No breakdown
May 4 1994 02:06:11:340PM	940504012	Steve Wolfe	Shot # 940504012 Load Br predictor using average in place of F25 Still using -3e-3 offset, probably wrong. ON previous shot there seemed to be a very serious difference between the flux plots using F25 and those using the average. Result: Still no breakdown, I apparently didn't put in what I thought I did! OH2L went to it's current limit! But the values look right in the predictor. The factor was wrong, should have been reduced from 40 to 25, but wasn't
May 4 1994 02:16:26:830PM	940504013	Steve Wolfe	Shot # 940504013 Fix factor on Br to 25. Result: Still no breakdown. but null looks better. Still some Br and a fair bit of Bz.
May 4 1994 02:38:57:030PM	940504014	Steve Wolfe	Shot # 940504014 Change Br0 offset from -3mT to -2mT Result: No breakdown
May 4 1994 02:50:30:580PM	940504014	Steve Wolfe	Shot # 940504015 Make EF4 more negative by 30A Result: Still no breakdown.
May 4 1994 02:55:44:730PM	940504014	Steve Wolfe	Shot # 940504015 Make EF4 more negative by 30A Result: Still no breakdown.
May 4 1994 03:09:07:680PM	940504014	Steve Wolfe	Shot # 940504015 Put EF2 back into the BR0 controller at -/+ 5000. starting at -.35sec Result: Got the flash back!! Breakdown is late. There seems to be a delay between the halpha and zmeter?
May 4 1994 03:14:44:030PM	940504016	Steve Wolfe	Shot # 940504016 Put Br0 offset back to -3mT Result: No breakdown
May 4 1994 03:30:45:880PM	940504018	Steve Wolfe	Shot # 940504018 Somewhere in the last n shots I lost one. The one that did not break down was 17, the one where it broke down after adding EF2 to Br was 16! For 18 put Br0 offset back to -1mT Result: no breakdown flux plot looks worse (if you can believe it)
May 4 1994 03:56:53:770PM	940504019	Steve Wolfe	Shot # 940504019 Back to -2mT, should be same as 16 Result: got the flash back Breakdown is at 15msec. The null looks terrible there!
May 4 1994 04:23:47:430PM	940504020	Steve Wolfe	Shot # 940504020 Punt, reload #8 pressure at 1.0e-5 Result: TF crowbar late. No breakdown. Note that on shot 8 the difference between F25 and the average was 50mWb and on thsi one it's 100mWb.
May 4 1994 04:37:11:860PM	940504021	Steve Wolfe	Shot # 940504021 Reload 16 and go another -30A on EF4 Steve H, thinks that's the wrong way. Result: No breakdown
May 4 1994 04:49:36:530PM	940504022	Steve Wolfe	Shot # 940504022 Go 30 A the other way to 982. Steve was probably right. Result: No breakdown. The new BZ predictor claims Bz is zero or slightly positive everyehere
May 4 1994 05:56:34:040PM	940504023	Steve Wolfe	Shot # 940504023 Go down 15 a (-997.5) Result: GOt a flash and a crowbar. Breakdown at 9msec on Halpha, later on Zmeter.

Engineering Operator Comments
Shot	Time	Type	Status	Comment
1	09:26:33:940AM	Plasma	Ok	Dud. No power supply problems.
2	09:38:18:410AM	Plasma	Ok	Dud (= no flash). INIT hung in FUELING tree. No other problems.
3	10:04:06:890AM	Plasma	Ok	Fizzle. Held in INIT 1 minute by TORVAC.
4	10:26:43:510AM	Plasma	Ok	Fizzle. OH1 PS comm fault at +170 msec; did not affect plasma. No othe
5	10:54:40:860AM	Plasma	Ok	Dud. No problems on this side of the aisle.
6	11:08:20:150AM	Plasma	Ok	Dud. No power supply problems.
7	11:29:36:400AM	Plasma	Ok	Flash. EF1L may not be responding to hybrid commands at t=-0.1 sec
8	12:34:38:890PM	Plasma	Ok	Flash; EF1 now looks fine though we didn't fix anything
9	01:01:55:270PM	Test	Ok	ok no plasma
10	01:16:59:910PM	Plasma	Ok
11	01:30:10:880PM	Plasma	Ok
12	01:47:44:120PM	Plasma	Ok
13	02:07:01:460PM	Plasma	Bad
14	02:19:41:580PM	Plasma	Ok
15	02:40:41:740PM	Plasma	Ok
16	03:01:23:650PM	Plasma	Ok
17	03:12:43:080PM	Plasma	Ok
18	03:25:38:270PM	Plasma	Ok
19	03:41:07:560PM	Plasma	Ok
20	03:58:29:800PM	Plasma	Bad	tf com fail
21	04:32:49:350PM	Plasma	Ok
22	04:43:28:740PM	Plasma	Ok
23	04:58:06:100PM	Plasma	Bad	com fail tf