Alcator C-Mod Run 951114 Information

Back to runs

Miniproposals
Miniproposal:120
Date Filed:10/18/1995
Title:Restart C-MOD 1995
First Author:Steve Wolfe
Session Leader:Steve Wolfe (shots 1-32)

Operators
Session leader(s):Steve Wolfe
Physics operator(s):Ian Hutchinson
Engineering operator(s):Vinny Bertolino,Joe Daigle

Engineering Operator Run Comment
120 Startup

Session Leader Plans

Physics Operators Plans

Session Leader Summaries
Entered: Jul 7 2004 03:49:13:960PM
Author: To Be Determined
Session Leader's Summary for 951114

MP#120 Startup 95
SL: wolfe
PO: Hutch
EO: Daigle/Bertolino

Goal was to finish up the startup MP:
Raise current to nominal levels (1MA)
Try again on pulse gas startup
Establish 800kA fiducial shot
Tune up of EF4 dead-time

The linear gas valve driver has been installed on A-side.
EF4 is on the fizzle detector
Started from a reload of 951103015, with changes to the pulse gas waveforms.

Although the goals were not fully met, enough was accomplished to declare
this MP complete. The remaining pieces will hopefully be dealt with in the
other runs this week.

Briefly, the current was only raised to 800kA, not 1MA. Shot 95111431 seems to
be a reasonable candidate for an 800kA fiducial shot. The pulse gas startup,
using the B-top valve, works, although the bugs aren't all out of it yet. The
pressure does seem to be reproducible, but the way we were using it, with the
pressure still rising sharply at t=0, may not be ideal. No work was done on
the EF4 deadtime.

The pulse gas startup was accomplished by a single short pulse around -0.1sec
on B-top (Pulse_gas_4), which had a plenum pressure of 20psi. The remainder of
the gas was supplied by "a-side" (actually located on k-side) which had
Willy's duty cycle driver on it, and was also filled to 20psi. The pressure at
t=0 was varied, by varying the length of the PG4 pulse, from around 8e-6 to
1.3e-5. No runaway discharges were produced, and most startups looked to me as
if they were being overgassed. With no plasma, a pulse getting 1.3e-5 at t=0
(shot #5) ended up with 4e-5 at t=0.4 sec.

The linear (duty-cycle) gas valve driver seemed to work pretty well, although
there was no obvious advantage over the direct drive version. It appeared that
we were not getting enough gas to push much higher than 1e20 without adding
gas from another valve. Of course, using B-top would be straightforward since
we are already using it for startup. For now, we decided to leave the
duty-cycle driver in; it can be changed back quickly if desired.

For a detailed blow-by-blow account of the run, see the Physics Operator's
Summary and the SESSION_LEADER topic in the logbook. Briefly:
Shot 1-6 Gas tests (shot 2 was Tuesday test as well)
Shot 7-9 Fizzles
Shot 10-18 600ka plasmas, and some fizzles
Shot 19-22 800kA attempts, limited to <700kA by the EF3 problem
described below; also some fizzles
Shot 23-26 EF3 tests
Shot 27-32 800kA shots (and fizzles)

One major problem encountered today is that the EF3 current regulation was
coming on at much too low a current. Apparently something has changed in this
circuit since last June. The symptom was that the Ip and RCUR were not
following their programs, and the EF3 demand voltage was rising and the output
not responding. After it was determined this was indeed the problem, the EF3
PLC limit was raised to 20kA, which allowed the current to get to about 8kA!
The engineers will try to have this fixed by tomorrow. In retrospect, this
may explain why the RCUR position and current were bad (current <600ka) on
951031 when Granetz was running, and better (current >600kA) on 951103015 with
the EF4 double swinging.

The dreaded longitudinal parity errors surfaced again, this time on the CII
and related signals. Joe Bosco got rid of them by once again jumpering out the
Probe crates. We also had some other CAMAC problems, which Josh managed to
repair and restore most of the lost data; these were apparently a
"communication" problem(?).

A minor problem was noted with the chopper. On a number of fizzles, the
chopper IOC'd after one on pulse. This was "fixed" by a change in the gain
programming as of shot 22. Note that the necessary change is in segment 2, as
the problem occurs at 0.11 sec.

There were also some problems with the alternator, resulting in a few shots on
which the DC breaker opened during check state, requiring us to go back to
recool and start again. This was apparently fixed by reducing the alternator
voltage.

Startup was very problematic all day, with the typical symptom being poor
burnout and current hesitation. The CII signals were typically around 0.02 to
0.03 mw/cm2/ster. Some field tweaking was done, as well as a lot of tweaks to
the initial gas. The last few shots actually had poorer current rises than
earlier in the day, although breakdown was slightly earlier. We ended up with
11 plasmas and 11 fizzles, not counting gas and EF3 tests.


Physics Operator Summaries
Entered: Jul 7 2004 04:36:55:190PM
Author: To Be Determined
Physop Summary 951114

PO: Hutch, SL:Wolfe
EO: Daigle, Bertolino.

Continuation of Restart MP120

Summary
_______

Established breakdown using the B-TOP valve for pulse fill.
Shot 22 and following give OK breakdown, although carbon is high.
Later shots used slightly lower fill. May be better.
In general the machine behaved as rather dirty: high CII, tendency to
alternate fizzles.

Established density feedback control using the "linear" chopped controller
on A-side valve. This works fine except there is a long delay to the first
chop if the demand is low. Seems to be an offset of about 10%.

Demonstrated a good 600kA plasma: Shot 18.

Discovered an EF3 PLC limit problem in trying to go to higher plasma current.
The actual EF3 limit is 0.43 times the programmed PLC limit. This is a linear
calibration problem. Eventually finnessed by raising the set point to 20kA.

Obtained a good 800kA plasma at 1.e20. Shape reasonable. Suggest shot 31 or
32 to start from.These could be fiducials but have the linear controller.
We are running out of gas to get much higher density.


Statistics
__________
Shot,Duration(s),Ipmax(MA) Shot,Duration(s),Ipmax
1 0.000 0.003 2 0.000 0.010
3 0.000 0.003 4 0.000 0.003
5 0.000 0.003 6 0.000 0.003
7 0.000 0.016 8 0.000 0.016
9 0.000 0.019 10 0.612 0.691
11 0.000 0.019 12 0.000 0.021
13 1.420 0.652 14 0.000 0.021
15 1.400 0.647 16 0.000 0.024
17 1.285 0.652 18 1.529 0.652
19 0.983 0.707 20 0.000 0.021
21 0.000 0.022 22 0.933 0.720
23 0.000 0.008 24 0.000 0.008
25 0.000 0.008 26 0.000 0.008
27 0.000 0.019 28 1.485 0.833
29 0.000 0.022 30 1.477 0.833
31 1.472 0.833 32 1.224 0.833

Finished summary for 951114 Total of 32 shots.
32 shots. 11 plasmas. 9 fizzles. 12 duds.
Fizzle = gt 17 lt 50kA.


Details
_______
Loaded 951103015
Pulse gas 4 B-Top. 100V, -.1 to -.092s.
Pulse gas 3 A-side really on K-side.
"Linear" chopping circuit controlled. Programmed.
Both at 20 psi. No prefill.

Alternator problem. Breaker pulled.
1 Gas Test B_Top opens all right and gives about 1e-5 at t=0. A-side
voltage chops at about 100Hz but with a 30ms delay.
2 Power supplies also. B-Top gas puff looks reproducible.
p(0)=10**(-5.06)

Add 1 ms to B-puff. (-.100to -.091) Set a downward ramp on A-side
100V to 10V at 0.13s.
3 Gas test. p(0)=10**(-5.02). RGA looks like a big air puff.

PG3 to (-.100 to -.089) PG4 Ramp 100(0) 26(30ms).
4 Gas p(0)=10**(-4.889)=1.3e-5 Still worried about the RGA
Ramp responds sensibly and opens quickly.

Disable A-side valve.
5 Gas RGA looks fine. p(0)=1.3e-5 quite reproducible.

Shorten initial A-side puff ramp to 20ms. Lower level to 14%.
Repurge A-side.
6 Gas p(0)=1.25e-5. A-side chops very spread out.

ZXL gain to 30 s^-1. A-side to 20% at 25ms.
7 Fiz 20kA, BD at 6ms. Density very low. Camac error, no carbon.
EF4 got -1055A

EF4 to -1115 from -1100A.
8 Fiz Got EF4 -1071A. BD at .5ms

Found that EF2U segment-1 controller is at .275 when the EF2L is at
0.37. So put EF2U to .37. Turn off probes to try to fix Camac.
9 Fiz 22kA, BD at 5ms. Fairly similar. Got CII trace, does not burn out.

PG4 fill pulse to (-115 to -106ms ) hence later on the pulse.
10 Pla Disrupts at .6s. 100Hz vertical instability.

ZXL gain back down to 4.
11 Fiz High Carbon level.

PG4 to (-115, -107) one ms shorter.
12 Fiz BD at 4.5ms, 25kA.

PG4 to (-115, -108) one ms shorter still.
13 Pla Went the duration. R_cur is at 65.5. This is too far in.
XL is not in the right position. It is too far in.
p(0)=9e-6, CII ~.01.

R_cur to .665 (from .655).
14 Fiz CII=.022. BD at 5ms (later).

EF4 to -1125A.
15 Pla Full length.

RXL to -.005 from -.01, ZXL gain to 8 from 4.
16 Fiz Nearly went but signs of possible runaway.

17 Pla Ramped down only to 400kA. Vertical instability in down ramp.
Data system errors, but fixed.

ZXL gain to 1 at 1s, and drawn rising up, to avoid vertical problem.
18 Pla Nice ramp-down. Slight hesitation.

Raise Ip to 800kA, NL to 1.e20.
19 Pla Disrupted at 1sec. Lots of Rcur and Ip problems staring at .36s.
Plasma leans out. Never gets to current.

Delay the Ip extra ramp till after .36s and make it more gentle.
20 Fiz

PG4 to (-135,-127ms). (20 ms earlier).
21 Fiz

PG4 to (-135,-127.5), 0.5 ms shorter. Also Chopper feedback lowered
near its turn-on to try to avoid trips.
22 Pla Again banging hard on the outboard limiter during the attempt to
raise the current. Tracked down the problem; EF3 is not following its
demand. Power supply tests.

Loaded from 951103013 PS test shot.
Programmed EF3 7kA and a +-2000V oscillation of the V.
23 EF3 Test. Got the 2000V pp sawtooth bu the current never went over 6kA.

PLC limit is at 15kA nominal, as per the day so far.
Pushed demand up to 8kA, and kept EF3V up for longer in the initial
upswing. There seems to be a factor of two difference between the
buss voltage and the apparent PS voltage.
24 EF3 Still limited to 6kA. The PLC is not letting us go beyond.

Lower PLC limit to 10kA.
25 EF3 Got even lower limit. About 4.3kA. This is essentially in constant
ratio with the set limit. This is a hard limit.

26 EF3 PLC controlled to 0.42 of what was set hence establishing the
calibration problem.

Reload shot 22 and try again with nominal EF3 limit at 20 (=8.5kA).
27 Fiz CII high, .027, vertical field slightly higher. BD at 5.3ms.

EF4 to -1125 (was -1115 but it should have been -1125, don't know why).
28 Pla Good shot. 7.5kA on EF3. Survived a big sharkfin at the end.
Shakey breakdown. CII peaks at .023.

Argon 20ms at .28s, RXL to -.002, NLo4 gain to 7 at .5s.
29 Fiz CII .025, 25kA.

PG4 to (-135,0128) (0.5ms less).
30 Pla

Tweaked the current and density traces to reach flat top a bit
earlier.
31 Pla Still hesitates in current rise. Current and voltage got up to
flat top earlier.

Set OH1 voltage to 100V from 130V at 5ms. Increased Ar puff to 30ms.
32 Pla Disrupts at 1.2s at the start of the sharkfin. Hirex got a signal.


Session Leader Comments
Nov 14 1995 08:09:34:100AMSteve WolfeMP#120 Startup
SL: wolfe
PO: hutch
EO: Daigle
Goal is to finish up the startup MP. Remaining tasks are:
Raise current to nominal levels (1MA)
Try again on pulse gas startup
Establish 800kA fiducial shot
Tune up of EF4 dead-time

The linear gas valve driver has been installed on A-side.
EF4 is on the fizzle detector

Nov 14 1995 09:03:57:920AM951114001Steve WolfeShot 951114001

Load 951103105.
Put pre-pulse PG4 (B-top) plenum at 20psi
Normal pulse on PG3 (also at 20psi) with linear driver
This will be a normal Tuesday test plus the gas puff.

Alternator fault. Gas test only.
B-top opened ok, got rising pressure at 9e-6 at zero, continues to rise
A-side doesn't look right.

Nov 14 1995 09:12:20:390AM951114002Steve WolfeShot 951114002

Tuesday test with gas.
B-top looks reproducible.
A-side still doesn't look right.
TCI_analysis was turned off; I turned it on for next shot.

Nov 14 1995 09:25:06:880AM951114003Steve WolfeShot 951114003

Put a ramp in PG3 programming.
Make pre-pulse on PG4 1msec wider.

Looks reasonable, but we are getting a burst of N2,O2, etc on RGA???

Nov 14 1995 09:34:57:510AM951114004Steve WolfeShot 951114004

One more gas puff test.
INcreasing initial PG4 puff by 2msec
Reduce PG3 puff.

Get 1.3e-5 at t=0.

Nov 14 1995 09:40:23:000AM951114005Steve WolfeShot 951114005

Test with a-side disabled.
No RGA junk;
Get 1.3e-5 at t=0, 4e-5 at t=.4

Nov 14 1995 09:59:21:310AM951114006Steve WolfeShot 951114006

Set the D2 regulator for 20psi, but it ended up higher. Never mind!
Re-purge a-side plenum
Ease off on A-side puff.
One more gas test.

1.26e-5 at t=0
A-side has too low a pulse?

Nov 14 1995 10:15:58:750AM951114007Steve WolfeShot 951114007

Try for plasma.
A-side puff program back up to 20%
Raise ZXL gain from 12/sec to 30/sec
ECE gate valve open

All power supplies ran.
Fizzle. bd at 6msec
No CII signal; store error on CAMAC.
May be too much Bz

Nov 14 1995 10:35:28:080AM951114008Steve WolfeShot 951114008

Make EF4 15A more negative.

Fizzle. Getting 1.7uohm on OH2U coax
No spectroscopy data this time.

Nov 14 1995 10:59:21:650AM951114009Steve WolfeShot 951114009

EF2U controller gain was at .247 instead of .37???
Do we still have ghosts in the PCS setup from the gain changes?
Set it back to .37 this time.
Turn off the Probes nodes and jumper out the probes crate to try to
get rid of LPE errors on the Spectr tree (CII signal)

Another fizzle.
Got the CII signal back; it definitely correlates with the probes crate.
I think we need to back down on the initial gas.

Nov 14 1995 11:15:45:300AM951114010Steve WolfeShot 951114010

Narrow the pg4 pulse and move it earlier by 15msec

Plasma- Disrupted in flattop.
Vertical instability.
Could be the x-point control

Nov 14 1995 11:23:07:570AM951114011Steve WolfeShot 951114011

PUt ZXL gain at 4 (hybrid units)
fizzle; CII signal is about doubled to .03

Nov 14 1995 11:39:42:040AM951114012Steve WolfeShot 951114012

Reduce the PG4 by 1msec in width
fizzle; PF crowbar
OH2U seems to go first.
Burnout still looks bad. CII down to .02

Nov 14 1995 11:58:34:640AM951114013Steve WolfeShot 951114013

Reduce the PG4 puff by another msec
plasma - looked like it made it all the way through.
Still a bad hesitation

Nov 14 1995 12:15:18:050PM951114014Steve WolfeShot 951114014

Move plasma out to .665 from .655
fizzle
PF crowbar - oh2u again
Chopper IOC fault

Nov 14 1995 12:31:53:650PM951114015Steve WolfeShot 951114015

Tweak EF4 a bit more negative (-1125 from -1115)
plasma, full length.
Breakdown looks better.
X-point still high and inside

Nov 14 1995 12:53:52:240PM951114016Steve WolfeShot 951114016

RXL out by 0.005 to -.005
ZXL gain to 30/sec
Alternator has dropped its voltage to 13.1 (when?)

fizzle. Bouncer at 14msec

Nov 14 1995 01:07:30:690PM951114017Steve WolfeShot 951114017

No change.

Looks good.
Data system errors. Looks like no spectroscopy and no hybrid data
Josh has recovered the data.
vertical instability in rampdown. ZXL gain is left on?

Nov 14 1995 01:16:50:970PM951114018Steve WolfeShot 951114018

Reduced ZXL gain in rampdown,and drew program to move up.

Looks good.
EF2L goes to zero early, EF4 swings back negative.
x-poimt still inside and high, but go to 800kA shot anyway.

Nov 14 1995 01:45:36:890PM951114019Steve WolfeShot 951114019

go to 800kA.
Raise nl_04 to 1.0e20

plasma - disrupted at eof.
Never got above 700kA
RCUR takes a big jump outside at around 400msec.

Nov 14 1995 01:58:26:010PM951114020Steve WolfeShot 951114020

Ask for gentler current rise.
fizzle - pf crowbar
Another chopper overcurrent.

Nov 14 1995 02:19:04:030PM951114021Steve WolfeShot 951114021

Move pg4 puff earlier by 20msec.
fizzle.
Another EFc overcurrent.
CII signal around .02 and no burnout.

Nov 14 1995 03:01:37:780PM951114022Steve WolfeShot 951114022

Adjust early gain on ZCUR3 to prevent chopper oc
Shorten PG4 puff by 0.5msec

plasma - disrupted around 920msec
EF3 acts like its on its current limit. PLC set point is 15kA, we
are runnning around 6000A, but it won't provide more voltage than around
500V.

Nov 14 1995 06:10:42:720PM951114026Steve WolfeShot 951114023-26

These were EF3 tests to demonstrate that the current regulator circuit
is turning on too early.

Nov 14 1995 04:08:32:350PM951114027Steve WolfeShot 951114027

Back to plasma - reload of shot 22
PLC limit set to 20kA on EF3

fizzle

Nov 14 1995 04:23:14:380PM951114028Steve WolfeShot 951114028

Tweak EF4 a bit more negative.
plasma - full length
got 800kA , RCUR looks good.
Density a bit soft. x-point a little high and inside

Nov 14 1995 04:31:59:410PM951114029Steve WolfeShot 951114029

Ar puff - 20msec wide
Move rxl 3mm out
Raise nl_04 gain to 7 at .5sec

fizzle -

Nov 14 1995 04:46:06:660PM951114030Steve WolfeShot 951114030

Reduce pg4 by 0.5msec
plasma - full length
density better but not great.
Early Current rise is very slow.

Nov 14 1995 04:54:14:070PM951114031Steve WolfeShot 951114031

Tweak current and density to bring up a littlke earlier

plasma - full length
startup looks poor, bad hesitation, only gts to 60kA by 30msec
density and current ok, maybe a little better than last time.

Nov 14 1995 05:00:39:210PM951114032Steve WolfeShot 951114032

Tweak OH1 voltage to 100V from 130V at 5msec
more Argon for JR

disrupted earlier.
early current rise looks worse.

Physics Operator Comments

Engineering Operator Comments
ShotTimeTypeStatusComment
108:48:53:090AMTestOk
209:02:55:360AMTestOk
309:17:54:620AMTestOk
409:29:20:130AMTestOkgas puff test.
509:35:38:660AMTestOkpulse gas test
609:54:07:320AMTestOkpulse gas test
710:07:01:920AMPlasmaOkNo faults
810:24:47:730AMPlasmaOk
910:53:25:320AMPlasmaOk
1011:05:45:700AMPlasmaOk
1111:19:53:030AMPlasmaOk
1211:31:57:860AMPlasmaOk
1311:48:27:680AMPlasmaOk
1412:05:37:570PMPlasmaOkoh2u fault, efc ioc
1512:21:46:810PMPlasmaOk
1612:41:59:430PMPlasmaOk
1712:54:17:350PMPlasmaOk
1801:09:30:060PMPlasmaOk
1901:28:26:760PMPlasmaOk
2001:54:49:430PMPlasmaOk
2102:14:22:370PMPlasmaOk
2202:29:26:850PMPlasmaOk
2303:34:56:070PMTestOk
2403:43:01:940PMTestOk
2503:48:28:670PMTestOk
2603:55:57:120PMTestOk
2704:03:55:360PMPlasmaOk
2804:13:42:740PMPlasmaOk
2904:27:55:840PMPlasmaOk
3004:37:29:170PMPlasmaOk
3104:48:02:650PMPlasmaOk
3204:56:05:220PMPlasmaOk