Miniproposals

Miniproposal: 147 Date Filed: 7/15/1996 Title: Enhanced Startup Techniques (Higher Pre-charge and Parallel EF4) First Author: Ian Hutchinson Session Leader: Steve Wolfe (shots 1-8)

Engineering Operator Run Comment

high precharge development, ef4s parralell,Hutch/Silva/Daigle

Session Leader Plans

Physics Operators Plans

Session Leader Summaries

Entered: Jul 7 2004 03:49:15:900PM

Author: To Be Determined

Session Leader Summary for 960816 MP#147 Enhanced Startup SL wolfe PO hutchinson EO Daigle The object of this run was to develop routine operation with Increased pre-charge and parallel EF4s. While not everything we wanted to do was completed, substantial progress was made. The run had to be stopped early because of a TF fault which blew multiple fuses and SCRs. However, we did demonstrate routine startup and reproducible plasma shots with the new configuration, and while some optimization should still be done, we are basically ready to run this way. In particular, there seems no reason to revert to series EF4 connection; the startup with high pre-charge seems at least as robust as the normal pre-charge, at least with RF assist. All shots today were run with breakdown at high-field (6T, ramping up) with subsequent ramp back down to 5.4T, obviously all under hybrid TF control. All successful shots had RF on at t=0, while two duds at the beginning of the day did not; however, it is not clear that the RF is really doing anything for us. We need to figure out how to startup at 5T as we used to do, but could live with this technique for a while if we had to. With the higher pre-charge, OH1 crossover happens around 400msec, which is already in flattop for 800kA; we see no problems around crossover. The programming still has some artifacts around 200msec which were put there to accommodate the original OH1 crossover. One problem with the high pre-charge is the tendency to start with slot divertor, then sweep out to the plates. The EF4 in parallel performed fine for all 8 shots today. Since an earlier run didn't start showing TF crowbar effects for 20 shots, we can't declare this problem fixed. However, neither the EF4 nor the TF snubber circuit which was put in to eliminate it seem to be implicated in today's TF fault. Joe shortened the EF4 crossover time in several increments, the last of which was not properly tested because shot 8 fizzled. Note that it is necessary to double the PLC current limits to get full current in the EF4s, since the PLC limit apply to the supply current. EF4_by_EF2 seems to be doing a good job of keeping the EF2 current down, and could probably be used even earlier. We still need to accumulate more constant shots for diagnostic and RF conditioning purposes. The RF got up to 500kW, according to the RF_NET trace. No heating was apparent on the ECE. The problem with the EF1L supply not following its demand, or following with an offset, was "fixed" prior to the run. The engineers caught it in the act of putting out a dc voltage on the control signal, and when they started to probe it, it went away. It worked fine for the entire run, but will probably recur sooner or later. Good plasmas were 5,6 and 7. Best shot was 7, although 6 ramped down a little farther. Shots 3 and 4 had mid-course disruptions after successful startups. Shot 3 had a marvelous vertical instability, one and a half cycles with the last swing recovering from +10cm. Scorecard: ---------- duds 2 fizzles 1 plasmas 5 ---------- total 8 All told, a pretty good result for a mere 5plasmas out of 8 shots.

Physics Operator Summaries

Entered: Jul 7 2004 04:36:56:490PM

Author: To Be Determined

Physop Summary 960816 Enhanced Startup MP147. PO Hutch. SL Wolfe, EO Daigle. Summary _______ A relatively encouraging run from the point of view of getting plasmas. Starting with the EF4s paralleled (but with the new bus sense snubbers) and higher precharge. A plasma was obtained on shot 3 after 2 duds. Thereafter five plasmas in a row led to a fizzle on shot 8 and a bad TF power-supply fault that terminated the run. We made good progress on the few shots that we had on getting stable and well controlled 0.8MA shots with higher precharge. OH cross-overs were no problem and we had begun to speed the EF4 crossover. Broadly, we have achieved the goals of the enhanced startup miniproposal although not (yet) at 5.3T. Also we did not establish what we did right compared with 960809. RF conditioning was beginning. Reasonably fiducial-looking shots (without much RF) were obtained. All shots today had 6 T at t=0. All the flashes had (1) 0.045 mtorr fill (2) RF at t=0. Use shot 7 as the basis for future plasmas. Tuning still needed: 1 Would be nice to get the startup to go at 5.3T 2 Need to change the EF4 program to a faster ramp rate. This will help to avoid the incipient EF2 saturation in the .2 to .4s time period. 3 Need to tune the current ramp-down. Rcur probably moves in too fast. 4 Need to tune the rxl (watching out for EF2 problem) to avoid the slot divertor in the .3s time period. Or else use it for physics studies. Plan was ________ ECDC overnight in **Helium** per request by Takase. Stop by 6am. Start with EF4s parallel. With new bus snubbers installed. Resistors set for the higher precharge. (.033, .006, .070 ohms) Crowbar OH at .12s, EF1 at .2s. Gas: standard pressures. Purge out any helium. We shall load from 960809003. But TF will be run with hybrid control, using the settings from 960815028. Plan is to continue the higher precharge development now that it is established that there is no problem getting a flash at the standard precharge settings. If necessary higher field at t=0 will be used and the field ramped down to 5.3T for the RF to condition. If we are extremely lucky, we will get constant shots for diagnostic calibration. Scorecard _________ Shot,Duration(s),Ipmax(MA) Shot,Duration(s),Ipmax 1 0.000 0.013 2 0.000 0.013 3 0.265 0.717 4 0.444 0.802 5 1.286 0.784 6 1.406 0.787 7 1.379 0.789 8 0.000 0.021 Finished summary for 960816 Total of 8 shots. 8 shots. 5 plasmas. 1 fizzles. 2 duds. Fizzle = gt 17 lt 50kA. Details ________ Coldstart. Start PCS Load segment 1 from 960809003. Adjust TF level to step down at 50ms, not 150. Copy TF from segment 2 to segment 1 Delay the step up in ZXL gain till .2s from .15s. PG 1 pulse width back to 20ms. (Argon). Gas puff to 40V in 40-80ms. Loaded n to model. Went to old PCS. Copied EF1L to EF1U. Loaded. Fill puff is 21ms. 1 Dud Read -9kA Ip. Got EF4 -1345, despite asking for -1405. A_out offset is 35A, which might explain it. Shot: 960809003 960816001 EF1L -6171. -6093. EF1U -6119. -6029. EF2L -2086. -2050. EF2U -2154. -2093. EF3L 121. 136. EF3U 102. 120. EF4L -1422. -1343. EF4U -1403. -1346. EFCL 48. 49. EFCU 7. 9. OH1 -20706. -20345. OH2L -26868. -26289. OH2U -27783. -27880. TF -158920. -169235. (get_bus_curs,[960809003,960816001] in user10:[hutch.mds] will do this for you) Draw EF4 to -1465. B_r to .005 from .006. 2 Dud Null looks very like those from 0809 that we did not understand. R_cur offset to 7000 in segment 1. Fill puff to 16ms. OH1 V to +100V from 0 at 1ms. OH2s to 0 from -25V. Ask for RF. 3 Pla Bdwn at .48s (for no obvious reason). 550kA at .1s. Disrupts at .26s because of a vertical instability that starts at .18s. EF2 rails at .215s. Density only 4e19. Gas seems to turn off. Found density was drawn low. Got RF early. Raise drawn density in early segment 2. Slow the increase of the RXL and ZXL gains 4 Pla Disrupts at .426s. Goes resistive at .4s 2pi bolo takes a big jump. Seems to be a moly injection, possibly RF. Density is still too low. Segment 2 rcur offset to 4000 from 8000. Ought to move the plasma out. by around 3mm. Put 100V gas puff at 40ms. Raise gas V in segment 2. Draw EF4 to 3500A positive from 2500. EF4 crossover tweaked by engineering 5 Pla Runs to 1.28 sec. Disrupts in ramp down. EF4 limited by PLC. Checked the all Ps; drawn stuff looks sensible. Draw RCUR 2mm further in (.669) OH1 V extended to 1.2s Increase gas at .07s. 6 Pla Disrupts at 1.4s. Argon to 30ms from 20ms. RXL programmed out to +.005 at .5s. 7 Pla Disrupt at 1.3s. EF4 ramps drawn faster. Crossover engineering twiddle. OH1 V demand to 160V, OH2 to 25V during flat top. Halved the rate of Rcur ramp during ramp down. 8 Fiz Got a more rapid EF4 ramp. but then, no plasma. TF trip. 47 fuses are blown. Big problems. Run ends because of the TF problems. Repair needed.

Session Leader Comments
Aug 16 1996 08:57:43:760AM		Steve Wolfe	MP147 Enhanced startup SL Wolfe PO hutch EO Silva ECDC overnight in **Helium** per request by Takase. Stop by 6am. Start with EF4s parallel. With new bus snubbers installed. Resistors set for the higher precharge.(.033, .006, .070 ohms) Crowbar OH at .12s, EF1 at .2s. Gas: standard pressures. Purge out any helium. We shall load from 960809003. But TF will be run with hybrid control, using the settings from 960815028. Plan is to continue the higher precharge development now that it is established that there is no problem getting a flash at the standard precharge settings. If necessary higher field at t=0 will be used and the field ramped down to 5.3T for the RF to condition. If we are extremely lucky, we will get constant shots for diagnostic calibration.
Aug 16 1996 10:10:38:780AM	960816001	Steve Wolfe	960816001: dud EF1L is following its demand. Various coils currents are low.
Aug 16 1996 10:27:12:000AM	960816002	Steve Wolfe	960816002: dud Null looks "good", just like on 960809
Aug 16 1996 10:49:59:440AM	960816003	Steve Wolfe	960816003: Tweak OH voltages; Change RCUR offset to 7000. RF on at t=0 lower fill to 15msec Plasma - early disruption, 250msec. vertical disruption. nl04 was set to 4e19. Why, Bob? vertical instability may be triggered by RXL gain switch, followed by ZXL gain switch.
Aug 16 1996 11:07:52:500AM	960816004	Steve Wolfe	960816004: Delay, slow down, RXL,ZXL gain switches. NL04 raised early. plasma, disruption . gets to 400msec. (OHcrossover) It starts as a slot divertor. Big gaps both sides. A moly injection precedes the disruption.
Aug 16 1996 11:39:49:120AM	960816005	Steve Wolfe	960816005: Move plasma out by changing seg2 offset to 4000 from 8000! Joe has shortened the Ef4 corossover time a little. INcrease IC_EF4 prog to 3.5kA. Push more gas in seg1 and 2. plasma disrupts at 1.25sec, just into rampdown. EF4 current limits at 1750 because of PLC set to 3.5kA RCUR and clearin start to slide at 1.0sec.
Aug 16 1996 11:53:46:700AM	960816006	Steve Wolfe	960816006:move rcur in by 2mm. Extend oh1voltage to 1.2sec. gas tweaks plasma-to 1.4sec, looks good.
Aug 16 1996 12:15:36:360PM	960816007	Steve Wolfe	960816007:move rxl out 5mm. More Ar for JR. Amanda is beginning to scan ECE. plasma to 1.35sec
Aug 16 1996 12:54:31:170PM	960816008	Steve Wolfe	960816008: tweak up ending (rcur). Still more Ar. More shortening of EF4 xover; ramps drawn faster. Modify feedforward voltages on OH1,OH2 Fizzle - all breakers pull; TF crowbar. bad noise from power room. >35 fuses blown in TF. EF4 went to zero current at 75msec (IHH's adjustment) and never crossed through???

Physics Operator Comments

Engineering Operator Comments
Shot	Time	Type	Status	Comment
1	09:46:57:090AM	Plasma	Ok
2	10:13:32:750AM	Plasma	Ok
3	10:31:28:240AM	Plasma	Ok	Comm Fault OH2U Bypass
4	10:52:24:710AM	Plasma	Ok	OH1 Self Powerd fired bypass
5	11:18:30:610AM	Plasma	Ok
6	11:44:22:280AM	Plasma	Ok
7	12:02:22:480PM	Plasma	Ok
8	12:23:10:670PM	Plasma	Ok	Loss Of many fuses in T,F.