Alcator C-Mod Run 970514 Information

Back to runs

Miniproposals
Miniproposal:170a
Date Filed: 5/13/1997
Title:Restart Operation: May 1997
First Author:Robert Granetz
Session Leader:Robert Granetz (shots 1-37)

Operators
Session leader(s):Robert Granetz
Physics operator(s):Steve Wolfe
Engineering operator(s):Vinny Bertolino,Frank Silva

Engineering Operator Run Comment
Restart Operation

Session Leader Plans

Physics Operators Plans

Session Leader Summaries

Physics Operator Summaries
Entered: Jul 7 2004 04:36:56:787PM
Author: To Be Determined
Physop summary for 970514

MP#170A
SL Granetz
PO Wolfe
EO Silva, Bertolino

Plan:
-----
This run will attempt to get the first C-Mod tokamak plasmas of 1997. If we
are successful in getting startup and current rise we will run cleanup
discharges at up to 0.8MA, with the shot length up to 1.2 sec (EOF).

We will start with integrated power system tests to verify OH and EF1
operation up to nominal high pre-charge values, then proceed to develop the
null to obtain breakdown and hopefully current rise. Specifically:

1) "Tuesday test": OH's only at 10kA, PLC control to check coax resistance
2) Engineering setup as on 960805015 (Low pre-charge case). TF on PLC control.
K-side to 20psi D2; B-top to 5psi D2- Enable both for hybrid.(why not?)
PCS copied from same shot. Just verify supplies all working together and
push on to higher pre-charge.
3) Engineering setup as on 960828005 (High pre-charge case). TF under PLC control.
K-side 20psi D2; B-top 5psi D2 (unchanged from above). Hybrid enable.
PCS copied from same shot, maybe adjust pre-puff to higher pressure.
4) Tweak up null to get a breakdown.
5) Tweak up current rise, using the usual black magic.
6) If it seems reasonable, back off the current to try to extend the pulse
length.
7) If we get full length pulses at lower current (say 600kA) then raise to
800kA for remainder of run.

Results:
-------
We got to somewhere between step 5 and 6. Two current rises were obtained, on
shots 8 and 26; shot 26 got to 600ka at 200msec. We also had numerous fizzles.
Both plasmas were extremely dirty and resistive. ECE measured temperatures
around 300eV.

The OH test on shot 1 indicated that we had a problem with the OH1 rogowski;
we also had CAMAC problems on the BUSLV signals on this shot. After lots of
testing and two more PLC test shots, it was found that the problem was related
to a loose connector housing. Shot 3 gave signal on all 3 OH rogowskis. Coax
resistances were fine on the tests.

Shot 4 was a low-precharge programing from #960805015. Everything worked
except the chopper. Gas pulse fired as programmed; this was the first hybrid
gas puff we've done this year.

Shot 5 was set up with high precharge from 960828005. Everything worked,
except no flash. Shot 6 was a Br tweak, which didn't get a breakdown. Shot 7
was a gas tweak, to higher pressure, and got a breakdown; first flash of the
year. Shot 8 was a repeat, and got a good current rise to over 400kA.

Shots 9 - 25 consisted of lots of fiddling, with some duds and some
fizzles. See the logbook (Physics_Operator topic) for details.

Shot 26 got another plasma, this one reaching 600kA at 200msec. The change
leading (?) to this plasma was an increase in the pre-puff to 26msec. It
seemed to like the higher pressure.

Shots 27 and 28 were basically recovery shots from the big gas load stirred up
by 26. Both fizzled.

Shot 29 was supposed to be an EF1L-only test shot to diagnose an oscillation
that appeared at shot 17 (see below). Unfortunately, we ended up blowing fuses
and incurred a long delay. Shots 30-35 were plasma attempts, only one of which
(33) even got a fizzle. These were also plagued by CAMAC errors on magnetics
signals (and others).

35 was the last real shot (a dud). JAS took two CAMAC-only test shots after
17:00 to try to track down the highway problem, but these did not experience
any difficulties.

Problems:
--------

Power supplies worked well, except for a chopper problem on shot 4 (blown
fuses again), and the EF1L problem that appeared on shot 17 and continued for
the rest of the run. This was the same problem that appeared earlier in the
week and then vanished, a larger than normal ripple on the voltage which
produced about a 100A ripple on the current. After the run the engineers found
a fault on the cosine board and replaced it.

In addition to blowing fuses, the test shot 29 produced collateral damage in
that the bus diagnostics on EF1L got zapped during the repair and testing
(hipot).

The null was somewhat variable for constant supply programming. It may be
useful to put higher gain, or integral gain, on some supplies, or to put Bz
feedback on.

We had several CAMAC highway problems throughout the day, including LPE's and
MEDOFL's, particularly on shot 1 and shots 30-35.

On a couple of shots Jim Irby noted possible fluorocarbon bursts. It is not
clear if these are causing any problems, and we still don't know where they
are coming from.

shot 27 had a visible flash on the TV, but no signal on the standard visible
light detectors. This may have been a weird CAMAC problem?

Comments:
---------
I did some tweaking of the RCUR offset in segment 1, but ended up back at the
original setting (7000). Changing this the first time required recalling the
predictor algorithm to incorporate the switched BP12 coil, and this seemed to
make some difference in the output. No other predictors were updated. This
should probably be done.

We clearly need a lot more cleanup, based on the two plasmas we got today. It
looks like the breakdown and even current rise favor high pre-puffs.

Cooling worked well all day. We had no apparent vacuum problems except the
known small leak and the fluorocarbon bursts.

Offsets are remarkably similar to last year.

Scorecard
--------
tests 5
duds 16
fizzles 12
plasmas 2
------ --
total 35

Session Leader Comments

Physics Operator Comments
May 14 1997 07:09:24:320AMSteve WolfeRun Plan for Wed. 970514

MP#170A Restart Operation: May 1997
SL: Granetz
PO: wolfe
EO: Bertolino,Silva

This run will attempt to get the first C-Mod tokamak plasmas of 1997. If we
are successful in getting startup and current rise we will run cleanup
discharges at up to 0.8MA, with the shot length up to 1.2 sec (EOF).

May 14 1997 07:09:54:350AMSteve Wolfe We will start with integrated power system tests to verify OH and EF1
operation up to nominal high pre-charge values, then proceed to develop the
null to obtain breakdown and hopefully current rise. Specifically:

1) "Tuesday test": OH's only at 10kA, PLC control to check coax resistance
2) Engineering setup as on 960805015 (Low pre-charge case). TF on PLC control.
K-side to 20psi D2; B-top to 5psi D2- Enable both for hybrid.(why not?)
PCS copied from same shot. Just verify supplies all working together and
push on to higher pre-charge.
3) Engineering setup as on 960828005 (High pre-charge case). TF under PLC control.
K-side 20psi D2; B-top 5psi D2 (unchanged from above). Hybrid enable.
PCS copied from same shot, maybe adjust pre-puff to higher pressure.
4) Tweak up null to get a breakdown.
5) Tweak up current rise, using the usual black magic.
6) If it seems reasonable, back off the current to try to extend the pulse
length.
7) If we get full length pulses at lower current (say 600kA) then raise to
800kA for remainder of run.

May 14 1997 07:27:43:180AMSteve WolfeIrby reports that t here was fluorocarbon generation last night when the ECDC ran down to
low pressure.

07:15 Do a coldstart = Looks ok.

PCS load from 960805015 - note that it took me three tries to get it typed right,
and PCS bombed out to the IDL prompt on each error; this seems to be a bug in PCS- should
fail to load gracefully.
Note that this shot has a 17msec pre-puff; should be fine (we're only using this
as an integration test anyway.

I got the acknowledge widget on the change to BP12 made last night. Don't bother recalculating
predictors for this test. We'll see when we get to the final setting
Load at 14-May-1997 07:27:12.00
May 14 1997 08:34:30:830AM970514001Steve WolfeShot 1 : "Tuesday" OH test
Irby found last night that there was a baseline drift problem with the TF rogowski.
As of now, it is still not fixed; use the BTOR value.

No bus rogowski traces. Jim says coax resistances look ok, assumng it's 10kA.
CAMAC store error on BUSLV.
Hybrid inputs show signal on OH2's, but not OH1. Also, live scope did not show OH1.
So we have two problems - Jim is doing CAMAC test; we need to check out the Rog. separately.
May 14 1997 09:05:53:940AM970514001Steve WolfeShot 1: Note that Tom "restored" data from BUSLV CAMAC after Irby had taken a test shot, so
the rogowski signals in the tree on shot 1 show a sine wave on OH1.
May 14 1997 09:18:06:990AM970514001Steve WolfeShot 1 : Rogowski checks out ok. Sine wave signal from sig. gen. shows up on BUSLV and
on Hybrid test shots (hybrid test shot is 272). Rogowski resistance is 9.7 ohms.

Hook it back up and try again test again.

May 14 1997 09:04:29:960AM970514002Steve WolfeShot 2: Try OH PLC test again.
Got the BUSLV camac, still no OH1 rog signal on anything.
May 14 1997 09:24:38:640AM970514003Steve WolfeShot 3: Can't find anything wrong with Rogowski, except housing on connector was loose?
Try another OH PLC test shot.

This time we have signal from all OH rogowskis. I guess tightening the housing fixed it.
On with the program, till the next SNAFU.
May 14 1997 10:02:46:540AM970514004Steve WolfeShot 4: Setup as for 960805015, go for everything.
Note that we are not getting TORVAC.GAS.PVALVE_*:Waveform signals. DOn't know why.

No Chopper, got everything else. No flash (Big Surprise!)
GOt gas - f.d. doesn't seem to shut it down until .2 sec.
Still no signals from PVALVE, but J. Goetz recovered the raw data.
May 14 1997 10:22:00:030AM970514005Steve WolfeShot 5: On to step 3. LOad from 960828005
For now don't bother with updating the predictors.
Load at 14-May-1997 10:11:29.00
They found some blown fuses in the chopper again.

No flash. Got all supplies. Null doesn't look terrible.
A little negative on BR, maybe a little low on Bz.
May 14 1997 10:42:42:070AM970514006Steve WolfeShot 6:
Change Br offset from .0065 to .007
Change pre-puff from 16 msec to 20msec.

No flash. Got more Bz this time, looks like EF4 current is 15A lower.
May 14 1997 10:52:14:350AM970514007Steve WolfeShot 7:
Raise pre-puff to 22msec. No change to fields, see how reproducible they are.

Got a flash. 30kA weak fizzle.
Field null is changed again.
breakdown at 7msec.
May 14 1997 11:05:45:180AM970514008Steve WolfeShot 8: repeat
flash - 400kA then disrupts at 140msec.
Dirty resistive but not bad for the first current rise.
May 14 1997 11:20:13:810AM970514009Steve WolfeShot 9:repeat
Dud. Null looks acceptable; just junk I guess.
May 14 1997 11:27:04:870AM970514010Steve WolfeShot 10: repeat
fizzle. Late breakdown 13msec.
Null a little different again.
May 14 1997 11:35:31:420AM970514011Steve Wolfeshot 11: repeat
dud. null looks ok. Try again.
May 14 1997 11:45:18:780AM970514012Steve Wolfeshot 12 repeat
dud.
May 14 1997 12:03:33:100PM970514013Steve Wolfeshot 13: raise Br to .0075 (up by .5mT)
fizzle. Late breakdown (13msec) but got a little current (45kA).
Looks like I went the wrong way on the Br tweak, as far as the flux plots are concerned.
May 14 1997 12:10:42:290PM970514014Steve Wolfeshot 14: lower Br to .0065 (down by 1mT)
dud - radial field looks ok, but now we have different vertical field.
May 14 1997 12:22:53:870PM970514015Steve Wolfeshot 15: repeat
fizzle 6.5msec bd. seems to be moving inside.
May 14 1997 12:37:44:500PM970514016Steve Wolfeshot 16: repeat
fizzle - late breakdown (15msec) and it bounced once moving out.
May 14 1997 12:49:26:790PM970514017Steve Wolfeshot 17: change rcur offset from 7000 to 6000. in seg. 1
Make predictor failed because of new BP loop.
Call the alg. then put in 6000. for the offset.

dud. RCUR went the way I expected, but maybe not the way the plasma wanted.
Null looks pretty bad. Try again before reverting.
May 14 1997 12:56:56:890PM970514018Steve Wolfeshot 18: repeat
dud. ok,untweak the rcur offset.
May 14 1997 01:11:25:990PM970514019Steve Wolfeshot 19: Set RCUR offset in Seg 1 back to 7000 (from 6000).
Note that this leaves the new predictor in place on Seg1 RCUR.

Vinnie reports that EF1L is back to having a lot of voltage hash on shot 18.
Actually it was bad on 17 as well, though not as bad.

fizzle. bd at 7msec, no current rise.
Rcur now thinks it's outside by 2000 a-m.

EF1L is still doing it's thing.
May 14 1997 01:22:23:320PM970514020Steve Wolfeshot 20: Change RCUR offset to 8000 (from 7000) in seg 1
fizzle. weak late bd (13msec).
RCUR still outside, a little further?
May 14 1997 01:32:03:500PM970514021Steve Wolfeshot 21: Change RCUR offset to 5000 (from 8000) in seg 1
dud. RCUR starts at 1000, compared to >2000.
Bz is less positive after 10msec.
May 14 1997 01:46:32:020PM970514022Steve Wolfeshot 22: Change RCUR offset to 7000 (from 5000) in seg 1
Vinnie is going to take out the reverse bridge on EF1L for this shot to try to diagnose
the big oscillations. We are seeing 100A ripple on the EF1L current.

late fizzle (15msec) and it moves in. 30kA current.
EF1L still bad oscillation.
May 14 1997 01:57:48:340PM970514023Steve Wolfeshot 23: Lower the pre-puff from 22msec to 18msec
Reverse bridge still out on EF1L, but added instrumentation.

fizzle. 11msec bd. current to 30kA, moves in.
May 14 1997 02:08:26:130PM970514024Steve Wolfeshot 24: repeat
dud.
May 14 1997 02:14:35:700PM970514025Steve Wolfeshot 25: lower pre-puff to 16msec (RSG made me do it).
dud.
May 14 1997 02:27:24:350PM970514026Steve Wolfeshot 26: raise pre-puff to 26msec
plasma - 200msec gets to 600kA.
May 14 1997 02:41:11:010PM970514027Steve Wolfeshot 27: reduce Ip program to 600ka.
fizzle? Saw a flash on the TV, but no signal on halpha or z-meter.
May 14 1997 02:48:25:910PM970514028Steve Wolfeshot 28: repeat
fizzle. 6.8msec, no current.
May 14 1997 03:04:50:180PM970514029Steve Wolfeshot 29: EF1L PLC shot only
Vinnie is taking out half the EF1L (running as a 6-pulser) to isolate the bad whatever.

Test didn't work. Put things back.
May 14 1997 04:07:09:140PM970514030Steve Wolfeshot 30: repeat of 28. (except the whole EF1L supply will be back in, after replacing some
fuses blown on the test shot 29.)

dud
lots of CAMAC errors on magnetics.
Can't diagnose the null directly, but it looks like EF4 is at 1420A, which is more than we had been
getting.

Try again.
May 14 1997 04:35:30:040PM970514031Steve Wolfeshot 31: repeat of 28.
another dud?
Still getting CAMAC errors on magnetics!
May 14 1997 04:43:07:410PM970514032Steve Wolfeshot 32: try again. TWF claims to have fixed the camac

another dud. This time mflux gets data, although there is one mds error reading BP18_jk??
May be a radial field problem, but it doesn't look that bad.
May 14 1997 04:48:51:830PM970514033Steve Wolfeshot 33: drop pre-puff to 22msec

fizzle, bd at 5msec.
More camac errors on magnetics
May 14 1997 04:57:13:850PM970514034Steve Wolfeshot 34: drop pre-puff to 20msec
dud
May 14 1997 05:03:20:480PM970514035Steve Wolfeshot 35: rise pre-puff to 24msec
dud

Engineering Operator Comments
ShotTimeTypeStatusComment
108:23:23:020AMTestOk
208:59:21:360AMTestOk
309:21:32:890AMTestOk
409:50:20:060AMPlasmaOk
510:15:37:290AMPlasmaOk
610:29:30:510AMPlasmaOk
710:46:24:790AMPlasmaOk
810:56:52:610AMPlasmaOk
911:07:41:310AMPlasmaOk
1011:20:47:310AMPlasmaOk
1111:30:11:940AMPlasmaOk
1211:40:27:440AMPlasmaOk
1311:58:24:410AMPlasmaOk
1412:06:46:700PMPlasmaOk
1512:17:29:120PMPlasmaOk
1612:29:05:330PMPlasmaOk
1712:44:53:010PMPlasmaOk
1812:55:09:390PMPlasmaOk
1901:05:40:110PMPlasmaOk
2001:16:26:920PMPlasmaOk
2101:26:05:370PMPlasmaOk
2201:38:27:390PMPlasmaOk
2301:54:16:510PMPlasmaOk
2402:04:44:740PMPlasmaOk
2502:12:47:190PMPlasmaOk
2602:21:55:260PMPlasmaOk
2702:33:16:990PMPlasmaOk
2802:43:35:910PMPlasmaOk
2903:01:17:650PMPlasmaOk
3004:00:25:530PMPlasmaOk
3104:24:48:990PMPlasmaOk
3204:36:26:070PMPlasmaOk
3304:44:59:600PMPlasmaOk
3404:54:33:550PMPlasmaOk
3505:01:50:830PMPlasmaOk
3605:47:35:200PMTestOkjosh test
3705:54:59:280PMTestOkjosh