Miniproposals

Miniproposal: 070 Date Filed: 11/28/1994 Title: Discharge Development; increase SN elongation First Author: Steve Horne Session Leader: Steve Horne (shots 1-10)

Engineering Operator Run Comment

higher elongation

Session Leader Plans

Physics Operators Plans

Session Leader Summaries

Entered: Jul 7 2004 03:49:11:757PM

Author: To Be Determined

run 940104 -- mp 70, discharge development; increase elongation. Horne, Snipes, Daigle. SUMMARY This run ended early (noon) due to loss of 13.8 (a Cambridge electric problem.) We accomplished one explicit goal (besides the implicit one of getting first 1995 plasma) -- the EFC series resistor has been reduced (for shot 10) to approx 60 milliohms, which should increase the speed of the EFC response. This shot had the oscillation (incipient instability) I wanted to use as a diagnostic. Shot 11 had some significant changes to the vertical control loop to improve stability -- fast zcur on wire 10, which has 10x usual D gain (PCS DOESN'T REFLECT THIS, SO BE CAREFUL!) but we lost the 13.8 before trying it. PCS can use 4 segments now, though I didn't get a chance to use more than 2 today. Shots 9 and 10 are nice diverted plasmas which could be (relatively) easily programmed as slot divertors, so I reccomend them for future runs this week (if any!). I tweaked the startup a bit to improve EF1L behavior -- segment 1, shots 9, 10 have the change. More work could be done here. I do not consider this run complete, since we were unable to test different P and Z gains on zcur, and did not try moving zcur up at all. Run plan and shot log follow. Steve PLAN -- Attempt higher elongation. Tricks to try: reduce EFC series resistor to get faster response. Wire 10 has 10x usual derivative gain. Put fast zcur there. Piggyback -- Develop a diverted rampdown to go with the limited rampdown from last run. Modification of wire 10 for higher derivative gain -- Changed the input cap from 10 nf to 100 nf = .1 uf; series r from 3.32 K to 320 Ohm. This change increases the gain by a factor of 10 across the full bandwidth. (The alternative; changing C but not R, rolls off the increase in gain at around 1 khz or so.) Setup -- use segment 2 from 941222011 (diverted shot). Segment # 1 941222033, segment 1 (Last shot's startup) 2 941222011, segment 2 (Last diverted shot) 3 " " turned off initially 4 941222033, segment 3 Limited rampdown; off IP wire, segment 3 has 0 IP drawn, gains like segment 4. Saved as user9:[horne]950104001 REMEMBER TO PROGRAM IP FLATTOP TO 600 KA! Put fast zcur on wire 10; set controller to zero. There seems to be a factor of 2 more chopper gain available in the controller. Why? Set it to zero. Object is to compare PID outputs to see effect of increased gain. SHOT LOG Plan -- get the first shot of the year; then change the chopper resistor. 1 OH test. The increased derivative gain on wire 10 is very evident -- 20 v p-p noise. Reduce D gain from 8 to .8 (equiv to old values 80 to 8). 2 dud. PF breaker (ef1 L, at commutation.) However the hybrid camac didn't trigger, which implies the jorway didn't fire. Since that controls the matrix switching, many bad things happened. Shot 1 seems to have been completely nominal, however. Set invert times to .2 (to protect system if something wierd happens) and try again. 3 PLASMA. Changing invert times must have fixed the problem. 4 Plasma. Nasty disruption at end of shot, probably because plasma flattop extended past 1 sec. Fix that, try again. Thereare other problems; may hgave to retreat back to a normal diverted shot, instead of this slot divertor configuration. 5 Very similar to 4. REDUCE CHOPPER SERIES RESISTOR FROM 150 TO 100 MOHM. The EF4 power supplies are giving us much closer to the programmed current. Program EF4 current to ramp down to 1500 A starting at .9 sec, to keep current in the EF2 coils. 6 EF4 single-phased starting at .5 or so. The radial position control pushed ef3 to the rail. Probable cause was an obscure ef2l power supply failure . Big changes for next shot, to get plasma shape under better control. Load 941220001, which has strike point predictors and controllers set up, but no gain. Reduce density to 8e19, IP to 600 kA, adjust EF4 programming. 7 Very nice diverted shot. Elongation 1.54-1.55. rcur in .5 cm, to raise elongation and provide an outer gap. Tweaks to zxl and xu for stability during rampdown. The zxl tweak will do nothing since P gaim is zero at 1 sec; fix for shot 9. 8 dud. The ef1l supply is still giving us more current than we want. Increase integral gain from 3 to 4; hold proportional gain on longer (segment 1) 9 Beautiful shot. DROP EFC RESISTOR FROM 100 TO 60 MOHMS. Reduced current limit to 2600 A. Program ZXL lower by 1 cm. 10 Nice shot. Elongation 1.6; incipient instability. Copied fast zcur to wire 10, to use increased P gain. Increased D from .8 to 1. Reduced P from 9 to 3. 13.8 KV LINE IN STREET HAS OPENED. RUN IS OVER!

Physics Operator Summaries

Entered: Jul 7 2004 04:36:52:800PM

Author: To Be Determined

The run proceeded very well with only minor glitches until the 13.8 kV line failed. Elongation only reached 1.66 since these were the preparatory shots before pushing the elongation. The last shot oscillated in Z position by about 1 cm throughout much of the shot. So, the fast ZCUR is still not correct. The run needs more time to continue to change gains. Shot,Duration(s),Ipmax(MA) Shot,Duration(s),Ipmax 1 0.000 0.008 2 0.000 0.011 3 0.239 0.499 4 1.097 0.611 5 1.093 0.608 6 1.035 0.613 7 1.464 0.636 8 0.000 0.011 9 1.489 0.639 10 1.497 0.642 Finished summary for 950104 Total of 10 shots. 10 shots. 7 plasmas. 0 fizzles. 3 duds.

Session Leader Comments

Physics Operator Comments
Jan 4 1995 08:32:18:830AM	950104001	Joseph Snipes	Discharge development run to increase the elongation for future RF studies. Horne also intends to "enhance" the chopper (EFC) capabilities by reducing its in series resistance to allow faster and larger chops up to 3 kA. Shot 1 OH1 test shot.
Jan 4 1995 08:42:20:500AM	950104002	Joseph Snipes	Shot 2 Attempt higher elongation. Prefill set to 1.4 x 10^-5 Torr. Wire 10 has 10x usual derivative gain. Put fast zcur there. Setup -- use segment 2 from 941222011 (diverted shot). 941222033, segment 1 (Last shot's startup) 941222011, segment 2 (Last diverted shot) " " turned off initially 941222033, segment 3 Limited rampdown; off IP wire, segment 3 has 0 IP drawn, gains like segment 4.
Jan 4 1995 08:46:14:940AM	950104002	Joseph Snipes	Shot 2 Hybrid errors. No plasma. TF and some OH supplies. It might be useful for diamagnetic loop and/or locked mode corrections.
Jan 4 1995 09:10:39:600AM	950104003	Joseph Snipes	Shot 3 We don't understand why the hybrid camac did not run, so we will change the invert time to 0.1 sec to avoid railing the power supplies if the hybrid does not work. Got plasma! Disrupted at 0.23 sec due to power supply inversion.
Jan 4 1995 09:25:54:840AM	950104004	Joseph Snipes	Shot 4 Changed inversion times back to their usual values. Decent plasma, but disrupted in the tail due to a jump in the density. Looking at EFIT, it is clear that the strike point starts moving up the nose on the inboard side. The X point moves up also.
Jan 4 1995 09:45:10:010AM	950104005	Joseph Snipes	Shot 5 Modified Ip waveform to begin rampdown at 1 sec. It had been somewhat later since the current of an 0.8 MA shot had been simply chopped off at 0.6 MA. Still disrupted at 1.1 sec.
Jan 4 1995 10:07:16:740AM	950104006	Joseph Snipes	Shot 6 Modified EF4 waveform to start ramping down at 0.9 sec instead of 1 sec to give more current to EF2 during the rampdown to try to maintain plasma control. Reduced EFC series resistor from 150 mOhm to 100 mOhm. Lost EF2L and EF4. EF3 also railed. Disrupted at 1.1 sec.
Jan 4 1995 10:44:06:030AM	950104007	Joseph Snipes	Shot 7 Reduced density to 8 x 10^19 m^-2. Modified EF4 programming so that it reached 2000 at about 0.35 sec instead of 0.255 sec. Loaded segment 2 from shot 941220001 to get strike point predictors and controllers set up with no gain. Good shot.
Jan 4 1995 11:15:15:400AM	950104008	Joseph Snipes	Shot 8 Moved RCUR from 0.657 m to 0.652 m to give a larger outer gap. Modified ZXL and ZXU late in the discharge to pull the X points out of the vessel during rampdown to improve vertical stability. Extended proportional gain on ZXL out to 1.6 sec. For the next shot, the ZXU proportional gain will also be extended out to 1.9 sec. Dud!
Jan 4 1995 11:23:20:060AM	950104009	Joseph Snipes	Shot 9 Raised prefill to 1.5 x 10^-5 Torr. Extended proportional gain on EF1L to near t=0. It was stepping down at -0.02 sec. Also, increased integral gain from 3 to 4 in segment 1. Good shot!
Jan 4 1995 11:42:31:940AM	950104010	Joseph Snipes	Shot 10 Dropped EFC resistor to 60 mOhm. Moved ZXL down about 1 cm throughout most of the shot. Good shot!
Jan 4 1995 12:06:49:860PM	950104010	Joseph Snipes	Shot 11 Changed fast ZCUR controller to wire 10, copying the programming from wire 3, which will give us 10 times higher derivative gain. So, the derivative gain was reduced to 1 and the proportional gain was reduced to 6. Increased current limit from 2600 to 2700 amps on EFC. There will be no shot 11 since the 13.8 kV power line coming to the lab blew up. End of run.
Jan 4 1995 01:19:56:480PM	950104010	Joseph Snipes	The run proceeded very well with only minor glitches until the 13.8 kV line failed. Elongation only reached 1.66 since these were the preparatory shots before pushing the elongation. The last shot oscillated in Z position by about 1 cm throughout much of the shot. So, the fast ZCUR is still not correct. The run needs more time to continue to change gains.

Engineering Operator Comments
Shot	Time	Type	Status	Comment
1	08:26:48:770AM	Test	Ok	OH coax test.
2	08:38:37:970AM	Plasma	Ok	hybrid faults
3	09:00:42:910AM	Plasma	Ok	no faults
4	09:12:28:170AM	Plasma	Ok	No Faults.
5	09:33:16:120AM	Plasma	Ok	No Faults.
6	09:49:42:670AM	Plasma	Bad	ef2lb and ef4 single phasing
7	10:41:22:660AM	Plasma	Ok	No Faults.
8	11:02:06:660AM	Plasma	Ok	No faults.
9	11:18:07:200AM	Plasma	Ok	No Faults.
10	11:37:35:490AM	Plasma	Ok	O.K. No Faults.