Alcator C-Mod Run 951027 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 1-16)

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

Engineering Operator Run Comment
day 2 plasma start up

Session Leader Plans

Physics Operators Plans

Session Leader Summaries
Entered: Jul 7 2004 03:49:13:943PM
Author: To Be Determined
Session Leader Summary for 951027
MP#120 C-MOD restart 1995
SL: Wolfe/Horne
PO: Horne/Wolfe
EO: Bertolino/Daigle

The plan for today's run was to get the power supplies fixed and then proceed
to make plasmas. After that we were supposed to tune up the plasma and a lot
of other ambitious things. As it turned out, we got plasma, left some supplies
in uncertain condition, and found a possible problem in the Hybrid. Progress
(toward standing by) was definitely made. Obviously, MP#120 will need at least
another day next Tuesday.

We started by trying to get the problem with the OH2's on inversion sorted
out. This was a very puzzling situation, as inverting from negative current in
either or both OH2 supplies caused fault indications in the bypass cabinet.
This did not show up last week, apparently because tests then were only done
up to 10kA and the problem didn't manifest until around 20kA. The first 8
shots were spent trying to isolate and diagnose this problem, without
noticeable success.

On shot 9, having ascertained that it was unlikely to endanger anything, we
set up for a full plasma attempt, on the theory that a flash would either
happen or not before the inversion problem occurred. There was also
speculation that the alternator loading might prevent the noise or whatever
from reaching the level to cause the fault. This turned out to work, in the
sense that the fault did not come up again all day. Note that it isn't fixed,
just hiding.

Shot 9 had two problems. The TMX supplies (EF2's, chopper bias) did not come
on because their contactor opened on an earlier access fault and was not
reclosed. More significantly, the EF4 had a shoot-through problem near
crossover and blew some more fuses. It was also noted that the EF4 is
apparently not listening to the fizzle detector. We decided to proceed by
disabling the reverse supply on EF4 after replacing the fuses. Gwinn and
company spent much of the remainder of the run figuring out a better crossover
circuit.

Shot 10 was the second plasma attempt, the first with no power system
problems. Amazingly, we got a 400kA , 150 msec plasma! The density reached
5e19 on NL04. We probably should have stopped at one-for-one, but we didn't.
It was noted that the demand signal to the chopper was wrong, consisting of
just the wavegen output and no PID output. This seems to have been the case
for the past week, at least.

Shot 11& 12 were a straight repeats, and fizzled.

On shot 13, we tried moving the fast ZCUR from wire 3 to wire 11, to try to
get the chopper demand back. The shot was another fizzle, and the M-chopper
demand was unchanged. This seemed to isolate the hybrid problem to a DAC card.

On shot 14, we dropped the pressure from 1.8 to 1.5e-5 and turned off the
chopper. We got back essentially the same 400kA shot, which again died
vertically at 150msec. At this point Ian convinced me we shouldn't continue
without trying to get the chopper working.

Horne changed out the suspect DAC card on the hybrid. Data system only tests
were then carried out (#15, 16) with the result that the same bad signal
appeared on the chopper demand line. At this point, we punted. After the run,
Horne checked out the hybrid hardware and found it to be functioning
correctly. As I write this, he is poking at the software and alternating
between cursing and exclamations of Eureka. He seems to have found a bug in
his new software.

Plasma attempts: 6
Plasmas: 2
Fizzles: 3
Duds: 1 (TMX supplies off)




Physics Operator Summaries
Entered: Jul 7 2004 04:36:55:113PM
Author: To Be Determined
Physop log, 951027 Horne

See Wolfe's SL log for details on power supply faults of various sorts

Good news is we Made A Plasma on first legitimate attempt (shot 10).
(Shot 9 was an attempt, but some power supplies didn't fire.)
Shot 14 was also a plasma.

I spent my time tracking down a peculiar problem with the hybrid output
to the chopper. Apparently there was no feedback; only
the preprogramming appeared at the output of the hybrid.
After some painful contortions, including changing out the DAC card which
was implicated, we punted the run. After some
more hardware testing the problem was diagnosed as software; the wrong gains
were being loaded. This tracked back to code I rewrote to support
the asynchronous segments; in the code which prepares the rt-node switch
tables, where I should have been counting M matrices, I counted the A
matrices (again). (cut, paste, forget to edit.)
Thus we got only the first 5 M matrices loaded into the DACs
(since there are 5 a-matrices) instead of the correct value (10).
This has been fixed, and a test shot shows (more) correct behavior.
I tender my most abject apologies for the screwup.

A point -- we shouldn't forget that the controllers weren't what we
expected today. If we have difficulties making a plasma next run,
we might look back to today's run and reconstruct (I can do that) the
controllers we really were running.

PCS note -- we've added a couple of buttons to each segment
which will recalculate all the predictors. We're also testing
built-in help on pcs -- it only works for these two new buttons so far,
but we'll probably add it in some other places.

Steve


















Session Leader Comments
Oct 27 1995 08:24:03:890AMSteve WolfeMP 120: Restart
SL: wolfe
PO: horne
EO: Bertolino/Daigle

Plan-
See yesterday's plan.
We'll start with OH2 non-commutation, double swing test. See if we can
determine what was wrong on last shot yesterday.

Tom reports that the highway errors got "fixed" by power cycling a
different crate. Highway was totally gronked this morning, but may be
ok now.

Oct 27 1995 09:07:57:520AM951027001Steve WolfeOH2 non-commutation under hybrid control, like yesterday #26 but not
commutate.

Joe says "same problem"
Hybrid asked for more current after t=0. Traced down to RCUR wire. This
would have been different on #26, as on that one OH1 was also on.

Engineers are out looking at power supply. I hope they aren't chasing
ghosts.

Oct 27 1995 10:05:03:540AM951027002Steve Wolfe951027002

PLC test OH2U only

Still getting a bypass fault at inversion.

Oct 27 1995 10:23:25:610AM951027003Steve Wolfe951027003

PLC OH2U test opposite direction current.
That seems ok

Oct 27 1995 10:28:09:340AM951027003Steve Wolfe951027003

About half of the full flux loops didn't get data on shot 3!*&^#$%^
It isn't fixed.

Oct 27 1995 10:56:53:330AM951027004Steve Wolfe951027004

OH2L PLC test, reverse current.
Joe Bosco has taken the PROBE crate off the highway. We'll see if that
helps.

OH2L crowbarred around -.5sec
Magnetics looks ok this time.

Oct 27 1995 11:23:02:350AM951027005Steve Wolfe951027005

Repeat of shot 4. They haven't a clue.

OK (no faults) that time. Big glitch on the current at -.27sec
Magnetics look ok

Oct 27 1995 12:04:12:830PM951027006Steve Wolfe951027006

OH2L negative current plc test
breaker pulled. Fault indicator in bypass

Oct 27 1995 12:13:41:890PM951027007Steve Wolfe951027007

OH1, OH2U&L PLC test. 10Ka

It worked.

Oct 27 1995 12:28:12:440PM951027008Steve Wolfe951027008

OH1, OH2U&L PLC test. 15kA on OH1, 20kA on OH2's
Breaker pulled, got a crowbar. OH2U.

Oct 27 1995 12:52:56:110PM951027009Steve Wolfe951027009

While looking for ghosts, go ahead and turn on everything else, and
try for plasma.

No flash. No EF2's. No chopper either.
EF4 had a problem at crossover.
No fault on the OH2's!!!
TMX problem due to access breach after contactors closed, not reclosed.

Oct 27 1995 02:04:38:870PM951027010Steve Wolfe951027010

Try again with TMX supplies on.
Disable the reverse EF4 supply.

Got a plasma. No faults. got to 400kA in 150msec. then disrupted.

Oct 27 1995 02:42:52:610PM951027011Steve Wolfe951027011

The chopper demand was not right on the last shot.
The current rounded over early, probably resistive.
Turn on the analysis tree this time.
Repeat.

No flash. There is an h-alpha signal, the TV filter must be in again.
fizzle.

Oct 27 1995 02:53:25:850PM951027012Steve Wolfe951027012

Repeat.
fizzle. 20kA

Oct 27 1995 03:06:02:740PM951027013Steve Wolfe951027013

Move fast zcur to wire 11. Chopper wasn't getting correct signal.

Another fizzle. about 20ka again
Chopper demand looks the same. It looks like only the Vout is coming
through.

Oct 27 1995 03:32:14:360PM951027014Steve Wolfe951027014

Drop fill to 1.5e-5
Turn off the chopper.

Plasma, got a PF breaker pull.
400kA. had some hesitation, but made it through.
OH2's look funny, as if driving the wrong way.

Oct 27 1995 03:50:02:350PM951027015Steve Wolfe951027015

Changed out a DAC card on the Hybrid.
Data system only test.

Error on INIT in Hybrid = A-matrix private bus not responding.
Chopper M-out is the same as before.

Oct 27 1995 06:01:26:750PM951027016Steve Wolfe951027016

Try the data system test again.
No change in chopper demand.

Punt, and work it out afterward.

Physics Operator Comments

Engineering Operator Comments
ShotTimeTypeStatusComment
108:46:17:760AMTestOk
209:58:36:050AMTestOk
310:21:04:920AMTestOk
410:53:02:790AMTestOk
511:19:15:360AMTestOk
611:49:46:550AMTestOk
712:11:46:760PMTestOk
812:16:17:950PMTestOk
912:45:22:170PMPlasmaBadno tmx's
1001:55:10:310PMPlasmaOk
1102:35:04:360PMPlasmaOk
1202:43:25:460PMPlasmaOk
1302:56:23:490PMPlasmaOk
1403:10:01:930PMPlasmaOk
1503:36:29:720PMTestOk
1603:51:57:780PMTestOk