| Miniproposals | ||||||||||
|
| Operators | |
| Session leader(s): | Jim Terry |
| Physics operator(s): | Steve Wolfe |
| Engineering operator(s): | Andy Pfeiffer,Ed Fitzgerald |
| Engineering Operator Run Comment |
| MP#458 (sort of) ELM structure, dynamics, and energetics SL: Terry PO: w |
| Session Leader Plans |
| Entered: Jul 18 2007 04:43:16:183PM |
| Author: Jim Terry |
| Study of ELM Structure, Dynamics and Energetics on Alcator C-Mod
This is a continuation of MPs #426 and #458. It will utilize the following new capabilities that should positively impact our investigation of ELMs: Cyropumping, A new array of probes mounted in the â??floorâ?? of the upper divertor, A new view and fast camera for Gas-Puff-Imaging just outboard of the typical location of the LSN X-pt., Additional fast magnetic sensors different toroidal locations, An inward shift of the radial location of the poloidal array of fast-diode views. In order to utilize the cryopump and the upper divertor probe array we will run USN plasmas with REVERSED field and current. In a well-boronized machine, we will begin by trying to produce Type 1 ELMs in the non-standard, so-called "JFT-2M" shape in the USN configuration. Will will exploit the cyropump to reduce the collisionality, attempting make the ELMs more reproducible and â??biggerâ??. After producing discharges with ELMs, we will study the ELM structure/dynamics - by examining the â??X-pt-GPIâ?? images (150,000 frames/s with a 1 microsec integration time) and what is seen on the poloidal array of fast-diode views; - by using the ASP in dwell mode (R=90.0, 91.0, and 92.0 cm mapped to midplane) to examine the timing (1 Mhz digitization rate) of filaments reaching the probe relative to those for the other fast ELM diagnostics; - by using the antenna loading measurements (250 kHz) to examine the timing of filaments reaching the antennas relative to those for the other fast diagnostics; - by continuing investigation of hi-freq magnetic oscillation at filament ejection, We will also study the ELM energetics - by performing power scan, concentrating on ELMs at high power (>3MW if possible), - by employing BES/DNB measurements for fast ne profiles during ELM crash, - by employing upper fixed probes in ion-saturation mode to make ELM timing and energetics measurements, - by examining the non-thermal electron generation at ELM crash using the ECE , - by examining soft x-ray signals at highest digitization frequency (100 kHz), - by employing ASP in dwell mode in the far-SOL (~90 cm) to yield an additional Te measurement in SOL (in addition to the ECE and TS measurements), - and by making ELMs at two significantly different values for the outer gap. We will start with discharges within the operational space where these discrete ELMs reproducibly occur, e.g. the USN version of shot 1060503022 (in the "JFT-2M" shape, with large upper triangularity (~0.79), small upper triangularity (~0.15),and low pedestal collisionality (target NL_04=5x1019 m-2)). We will increase the ICRF power from 2 MW to the highest available power, in order to examine the ELM frequency vs RF power. Once large ELMs are produced, shots will be repeated to get the best possible ELM measurements. We will then attempt to reduce the H-mode density (and collisionality) by using the cyropump. Diagnostics: Core and edge TS*, TCI*, ECE* (100kHz dig rate with outermost GPC channel in the pedestal and outermost GPC2 channels outside the sepx), midplane GPI*, X-pt GPI*, upper divertor probes*, Soft x-ray arrays (digitized at 100 kHz from 0.8 to 1.0 s)*, fast magnetics (digitized at 10 MHz from 0.7 to 1.2 s)*, ASP in dwell mode at R=90.0 cm, 91.0 cm, and 92.0 cm (mapped to midplane)*, upper divertor probes (digitized at 1 MHz from 0.7 to 1.2 s)*, IR camera* BES (viewing the pedestal with a 1 MHz digitization rate)* PCI (digitized at 10 MHz from 0.7 to 1.2 s), reflectometry, DALSA VB profiles *ed diagnostics are crucial Run Plan: 1. With well-boronized walls, start with USN version of shot 1060503022 (Ip=0.9 MA, ICRF power 2 MW, inner gap=2.0 cm, but with Bt=5.4 T and outer gap=2.0 cm) â?? observe discrete ELMs (4 good shots - with fast ELM diagnostics working). 2. Repeat with increased ICRF power - up to max available (2 good shots) 3. Repeat with cyropumping at optimum RF power level (2 good shots) 4. At optimum RF power and cyropumping (if that has improved the reproducibility of the ELMs) change the target density to NL_04=3e19 m-2 and NL_04=8e19 m-2 (2 shots per density â?? 4 shots total) 5.At optimum RF power, cyropumping, and target density, change the outer gap to 3.0 cm and 1.0 cm. (2 shots per gap â?? 4 shots total) 6.Repeat at optimum RF power, cyropumping, and target density with fast ELM diagnostics working (fast-magnetics, two GPI views, upper divertor probes, ASP in dwell mode, BES/NBI, PCI, X-ray arrays, antenna loading detection) (7 shots) 7. Repeat at optimum RF power, cyropumping, and target density with 6 of the fast-diode views switched to include the inboard radial array (2 shots) |
| Physics Operators Plans |
| Entered: Jul 18 2007 03:24:43:360PM |
| Author: Steve Wolfe |
Engineering Setup and Physop plan for Thursday, July 19, 2007 MP#458 (sort of) ELM structure, dynamics, and energetics SL: Terry PO: wolfe This run will be reverse field, upper null, with cryopump. No overnight ECDC, but do a 60C bake Two hours ECDC in D2, finishing by 8AM. Pressure 2e-4 Torr, sweep rmin/rdwell/rmax=44/45/103 tdwell/period = 20/60sec Run begins at 9:00 (lockup at 08:30) and ends at 17:00 Engineering setup: Power systems as on 1060503022, EXCEPT acoils configured (-Dtop +Dbot +Jtop -Jbot) enabled Gas setup: Fill B-Top with 6 psi D2 Hybrid enabled (PG4) fill B-side lower with 1 psi Ar Hybrid ENABLED (PG1) leave B-side upper as is Hybrid DISABLED (PG2) fill B-main (C-side) with 40 psi D2 Hybrid enabled (PG3) fill NINJA to 10 psi D2 and ENABLE. Enable the following gate valves and shutters, assuming no problems: ECE, VUV, DNB, Z-bolo shutters Torvac setup: Gate valves open after the shot ICRF: Transmitters 1,2,3,4 in RUN-ON by 8:30 Start with 2MW (D+E) from 0.7 to 1.2sec as on 1060503022 We will probably want to add J later in the day. LH: NO LH for this run DNB: yes (all DNB diagnostics) Cryopump: ready for operation after first successful plasma. See SL plan for list of required diagnostics. Physop plan: ------------- SL has requested upside-down JFT2-M shape (in reverse field). See logbook under Physics Operator for setup procedure done 1070719. The main idea is to get type I ELMs and then study them. Key parameters (per SL) are: Bt=+5.4 T Ip=+0.9 MA NL04=0.5e20 m-2 upper triangularity=0.79 lower triangularity=0.14 outer gap= 2.0 cm (this is different than 1060503022) inner gap=2.0 cm We'll be varying outer gap and density, and using the cryopump. I expect there will need to be some tweaking on the shape to get things lined up with edge TS, the upper divertor probes, the GPI, and to optimize the ICRF coupling. At some point we'll probably want to do a power scan to max available power. |
| Session Leader Summaries |
| Entered: Jul 19 2007 05:34:15:560PM |
| Author: Jim Terry |
| The run was plagued with difficulties, so we got essentially
none of the desired measurements. There were problems with power supplies that prevented us from running the highly-shaped "JFT2M" USN configuration with reversed field and current. We go one discharge that made it into ramp-down (1070719004). We got non-steady H-modes on shots 1070719002, 004, and 006. We may have gotten 3 ELMs on Shot 1070719006 at times 0.792 s, 0.819 s, and 0.875 s. As expected, the H-mode pedestals looked similar to those of the LSN, normal-field "JFT2N" shape. So it is probable that the desired plasmas can be produced, if the power supply issues are favorably resolved. In this configuration the outer strike point appeared to be on upper divertor probe # 6. Other diagnostics that were to be exploited for this run that worked as desired were: midplane GPI fast-diode arrays PCI upper divertor probes ASP in dwell mode B-side view angle camera view IR camera We didn't get the fast-magnetics at the 10 MHz sampling rate until after shot 006. We didn't see light with the X-pt GPI camera - reason as yet unknown. There was no light on the soft x-ray edge array - reason as yet unknown There was some signal on the fast BES channels, but it was very small. |
| Physics Operator Summaries |
| Entered: Jul 19 2007 06:14:47:047PM |
| Author: Steve Wolfe |
Physop summary for Thursday, July 19, 2007
MP#458: ELM energetics...
SL: Terry
PO: wolfe
EO: Pfeiffer, Fitzgerald, Cochran, Dekow
This was an unsuccesful run.
There were at least two problems with trying to reproduce an upside-down
JFT2-M shape in reverse field. The first, and the one that cost us most
of the day, was an instability in the EF1L power supply regulation
near crossover (zero current) which caused a burst of voltage spikes on
the bus, resulting in firing of the self-powered coil protection circuit,
and subsequent crowbar, leading to disruptions on shots 1, 5, and 8.
Thinking that this might be due to excessive voltage demand on the supply,
after shot 1 and again after the recurrence on 5 I reduced the gain on the
ZXL, which did reduce the voltage demand but did not cure the crowbars.
Power supply tests attempting to identify the cause of this problem
resulted in massive fuse loss on shot 11, and the remainder of the day was
spent recovering from this event.
The other problem, which is arguably more fundamental, was that the EF2U
supply was driven to zero current trying to get the upper triangularity
as programmed, while the EF2L was close to its positive current rail.
This situation was somewhat more extreme than had been found in the
original LSN normal field case, 1060503022, which had both EF2's close to
but not at their opposite current limits. The fact that these are at
opposite limits implies that the problem cannot be fixed by adjusting
IC_EF4 or EF2_by_EF4, both of which affect the EF2 currents symmetrically.
I had actually tried a small change in EF4 to try to rebalance the
EF2's and maybe get the EF2U off of zero for shot 8, as well as relaxing
RXL a little, but the shot disrupted early because of the EF1L supply problem.
Shots 2 and 6 disrupted between 0.9 and 1 sec with vertical problems
associated with the railing (zeroing) of EF2U.
The only full length plasma was shot#4, which survived, but had gaps
a little too small and upper triangularity a bit low. This shot also
had EF2U solidly at zero current for the entire flattop, but managed
to survive in spite of that. Jim thought there were two or three ELMS
(the point of the experiment) on shot 6, which disrupted around 1sec.
After the crowbar and disruption on shot #8 we began investigating
the cause of the EF1L problem. An EF1L double swing was programmed in
segment 1 (seg 2 off) and run on shots 9 and 10. Between these shots
the iso-amp on the bus voltage signal on EF1L was swapped with the one
on EF1U, because the former was too noisy to show useful data. This had
also beent the case on the plasma shots. On shot #10 there was a clear
burst of spikes above 600V on the EF1L bus before crossover, coincident
with the observed oscillation on the supply circulating currents. We had
previously noted that there was not nearly as much voltage spiking or
oscillation on the EF1U supply when it went through crossover. Bill therefore
tried swapping the slew boards from the two supplies and re-doing the
test. This resulted in 24 blown fuses in the supply. It turns out that
the two boards are not interchangeable, and putting the EF1U board in
EF1L resulted in shoot-throughs on all the devices. No SCRs were lost,
but the supply of spare fuses is now down to about 30.
The rest of the day was spent changing the fuses, hi potting the
SCRs and testing the supply. On shot #18 we re-did the PCS controlled
double-swing test on EF1L and confirmed that while the supply was
working again, it still had the same oscillation and bursty voltage as before.
Shot #19 repeated the same test on EF1U, and revealed that it had a similar
behavior, though perhaps not as severe. Re-examining the old LSN shot
1060503022 I found that there was in fact a burst on the voltage on EF1U
at about the same time in that shot, near EF1U crossover between .4 and .5
sec, which had not resulted in a crowbar but may have been close. At that time
there was also a noise problem on the EF1U bus voltage so it is not possible
to see how large the spikes on the voltage were. Looking at some other
plasma shots it seems that the size of this effect is related to the
rate of change near crossover, which implies that the peculiarities of the
this equilibrium, which does approach zero quite fast compared to more normal
shots. It is not clear whether anything can be done to reduce the resulting
unstable oscillation in the supply enough to run these cases without
compromising some other aspect of the control. Even if Bill can manage
to do that, the Ef2 problems may prevent us from successfully carrying out
this experiment unless we can back off on the shape somewhat, perhaps by
allowing larger lower triangularity, which would reduce the demands on
both EF2L and EF1L.
Scorecard:
Tests 11
Duds 0
Fizzles 2
Plasmas 6 (only one full length)
======== ==
Shots 19
NOTE TO NEXT PHYSOP: The shot currently loaded in PCS_MODEL is an EF1 test,
with Seg2 turned off and the EF1's in seg1 programmed to double swing.
Don't use this unless you are doing EF1 tests with the engineering
crew!
|
| Session Leader Comments | |||
| Jul 24 2007 10:30:29:533AM | Jim Terry | Setup of GPI diagnostics:
NINJA setup to puff D2 from caps K-bolo and A-B lim. Trigger at 0.7 s, duration=0.14 s X-pt view trigger at 0.8 s, 2 microsec exposure (30000 frames) for shot 1 the fast-diode data window was from 0.780 to 1.304 sec diode # trans fiber # view F/no 1 GPI #3 OUTER SOL 36 F/1.4 2 GPI #2 OUTER SOL 22 F/1.4 3 GPI #1 OUTER SOL 5 F/1.4 4 GPI #4 OUTER SOL 23 F/5.6 5 GPI #5 OUTER SOL 10 F/5.6 6 #39 OUTER SOL 9 F/5.6 7 #40 OUTER SOL 7 F/1.4 8 #41 OUTER SOL 11 F/8.0 9 #42 OUTER SOL 12 F/5.6 10 #43 OUTER SOL 13 F/8.0 11 #44 OUTER SOL 14 F/5.6 12 #45 OUTER SOL 6 F/1.4 13 #46 OUTER SOL 29 F/5.6 14 #47 OUTER SOL 31 F/1.4 15 #48 OUTER SOL 32 F/8.0 16 #49 OUTER SOL 33 F/4.0 17 #50 OUTER SOL 34 F/4.0 18 #51 OUTER SOL 35 F/4.0 19 #52 OUTER SOL 3 F/2.8 20 #20 OUTER SOL 37 F/4.0 21 #21 OUTER SOL 38 F/4.0 22 #22 OUTER SOL 39 F/4.0 23 #26 OUTER SOL 40 F/4.0 B-side camera view is setup with the SONY camera, an ND 1 filter in the lens coupling system, the camera lens is at largest the aperture and focussed at infinity. The lens on the end of the bundle had an F/16 aperture and was focused at infinity. IR camera set and calibrated with NUC=1, filter=3 - data for 150 frames | |
| Jul 19 2007 09:34:05:567AM | 1070719001 | Jim Terry | disrupted at 0.528 s
NINJA puffed 8 Torr-l B-side camera worked IR camera worked Phantom camera worked |
| Jul 19 2007 10:05:35:290AM | 1070719002 | Jim Terry | disrupted at 0.964 s
H-modes Possible ELM at 0.812 s NINJA puffed 8 Torr-l B-side camera failed IR camera worked Phantom camera worked |
| Jul 19 2007 10:21:06:823AM | 1070719003 | Jim Terry | fizzle
NINJA puffed 8 Torr-l B-side camera worked IR camera worked Phantom camera worked |
| Jul 19 2007 11:08:54:990AM | 1070719004 | Jim Terry | shot made it all the way with good JFT2M shape.
ratty H-modes no ELMs Brian reports that the outer strike appears to be on upper Div probe 6. NINJA puffed 8 Torr-l B-side camera worked IR camera worked Phantom camera worked |
| Jul 19 2007 11:46:48:220AM | 1070719005 | Jim Terry | disrupted at0.597 s
NINJA puffed 8 Torr-l B-side camera worked (frame 13 is a good registration frame) IR camera worked Phantom camera worked For this shot I changed most of the F/nos on the diodes: They were changed to: fast-diode data window was still from 0.780 to 1.304 sec diode # trans fiber # view F/no 1 GPI #3 OUTER SOL 36 F/1.4 2 GPI #2 OUTER SOL 22 F/1.4 3 GPI #1 OUTER SOL 5 F/1.4 4 GPI #4 OUTER SOL 23 F/4.0 5 GPI #5 OUTER SOL 10 F/4.0 6 #39 OUTER SOL 9 F/4.0 7 #40 OUTER SOL 7 F/1.4 8 #41 OUTER SOL 11 F/5.6 9 #42 OUTER SOL 12 F/4.0 10 #43 OUTER SOL 13 F/5.6 11 #44 OUTER SOL 14 F/4.0 12 #45 OUTER SOL 6 F/1.4 13 #46 OUTER SOL 29 F/5.6 14 #47 OUTER SOL 31 F/1.4 15 #48 OUTER SOL 32 F/1.4 16 #49 OUTER SOL 33 F/2.8 17 #50 OUTER SOL 34 F/4.0 18 #51 OUTER SOL 35 F/4.0 19 #52 OUTER SOL 3 F/1.4 20 #20 OUTER SOL 37 F/4.0 21 #21 OUTER SOL 38 F/2.8 22 #22 OUTER SOL 39 F/2.8 23 #26 OUTER SOL 40 F/2.0 note there was an error in the tree for the F/no for diode 15. Prior to this shot the tree had the diode 15 F/no as 8.0; in fact is was 1.4! I should up date the tree for previous shots. |
| Jul 19 2007 12:20:11:520PM | 1070719006 | Jim Terry | disrupted at0.947 s
H-modes, but possible ELMs at 0.792 s, 0.819 s, 0.875 s NINJA puffed 8 Torr-l B-side camera worked IR camera worked Phantom camera worked |
| Jul 19 2007 12:13:06:847PM | 1070719007 | Jim Terry | fizzle
for this shot I moved the NINJA trigger back to 0.6 s the fast diode trigger back to 0.68 s the Phantom camera trigger back to 0.72 s NINJA puffed 8 Torr-l B-side camera worked IR camera worked Phantom camera worked |
| Jul 19 2007 04:24:23:150PM | 1070719008 | Jim Terry | disrupted at 0.615 s
NINJA trigger is 0.6 s fast diode trigger is 0.68 s Phantom camera trigger is 0.72 s NINJA puffed 8 Torr-l B-side camera worked IR camera worked Phantom camera worked, but there appears to be no light at all, even thru the disruption of shot 006. So after this shot I've slowed the frame rate down to 90909 f/s and increased the exposure time to 8 microsec. |
| Physics Operator Comments | |||
| Jul 18 2007 12:29:39:470PM | Steve Wolfe |
Starting setup for 1070719
MP#458 (sort of)
SL: JT
SL wants to run reversed field, flipped, i.e. upper null, version of
1060503022 (with a few other gratuitous changes).
The active segment is seg2.
This shot was normal field and dates to before the f10,f11 loop failures,
so the predictors have to be recalled as well as all the field reversal
stuff.
Import seg2 from 1060503022 into model from before 1070718007.
Start at the top - change the seg2 title, seg2 on seg3 off
Do the predictors first:
Some of the fluxihh predictors were from 02, some from 00, haven't
a clue whether this shot was actually done right or not. Assume it
was, and that the updates will be correct.
Note that to do the upper null version, I have to interchange the
upper and lower xpoint waveforms, interchange (or modify) the gains,
change the clearin reference point. I think the controllers are symmetric
so I don't need to re-calculate those, but I'd better check that.
First things, first. Pretend we were going to run this shot in reverse
field and start on the flips and such. I've already re-called the B21_PROJ
predictor. Also checked the Ip predictor to make sure it didn't have an
offset, which it didn't.
All_P's: flip EF2_by_EF4, IC_EF4, Ip
V#01-16: flip OH1_VOLT, CHOPPER, OH2U_VOLT, OH2L_VOLT, EF1U_VOLT, EF1L_VOLT,
EF2_U_VOLT, EF2_L_VOLT, EF4_VOLT, EF3_VOLT, TF_CURRENT
Check that ASYNC_PROS | dpcs_fizzle still has source_shot=1060517007 ok
If I stopped now, I think it would work for a lower null. But NO...
Jim wants it upside down and backward.
Back to CLEARIN and change the ref point: keyword: ZXA=.394
and call it again. The controller is up-down symmetric so leave it alone.
Now for the other waveforms, on the ALL_P's screen
Flip zcur (both of it); apply
Interchange RXL, RXU (use wire 12 which has zero_12 as a parking place)
apply.
Interchange ZXL, ZXU; apply (just to have it saved, it's about to change)
Flip ZXL, ZXU; apply
Put the screen away, and recall the zero_12 (use zero_all) to get rid
of the last waveform.
Now I need to deal with the gains. I think I need to interchange the gains
from RXL, RXU and ZXL,ZXU. Back to edit_wire screens. Unfortunately,
the load gains button doesn't let me specify another shot, just wire and
seg. I'll copy_wire from RXL, ZXL into wires 10,11 (from 5,6) temporarily,
then zero them when I'm done.
Wire 10: Load_wire from seg2 wire 5;apply, acknowledge that it's not an Ip wire.
Wire 11: Load_wire from seg2 wire 6;apply, ""
Wire 5(RXL): Load_gains from seg2 wire 07(RXU); apply
Wire 6(ZXL): Load_gains from seg2 wire 08(ZXU); apply
Wire 7(RXU): Load_gains from seg2 wire 10(RXL copy); apply
Wire 8(ZXU): Load_gains from seg2 wire 11(ZXL copy); apply
Wire 10(RXL copy) : Zero all; apply
Wire 11(ZXL copy) : Zero_all; apply
I think that completes the basic setup, except for the gap change
Jim asked for. I need to do a sanity check on these waveforms, but
first do a build_all and see what breaks.
Now save it to /home/wolfe/pcs_trees/ shot#1070719666
RXU is at -.055m ! Can that be right? Import the original to seg3 and take
a look. Yes, that's where RXL is on that shot.
Note that EF2_by_EF4 is at +4000 (original was -4000) which seems uncomfortably
close to the rail, but that's how it was set up.
Maybe I have it setup ok. Back to the instructions.
Jim wants:
Bt=+5.4 T (changed from 5.65 in previous note)
Ip=+0.9 MA
NL04=0.5e20 m-2
upper triangularity=0.79
lower triangularity=0.14
outer gap= 2.0 cm (this is different than 1060503022)
inner gap=2.0 cm
The nl04 program is set to 4.2e19, I'll raise it to 5.0e19
Move rcur out by 2.5mm, although it will need adjustment (and I won't
be able to trust EFIT anyway). Make adjustments from .55sec to 1.5sec
(3 points); apply
Re-import seg3 from -1 (so I don't leave the unmodified seg in place)
Build_all
Save as /home/wolfe/pcs_trees shot#1070719666
Take a break and save this logbook entry.
| |
| Jul 19 2007 08:06:44:327AM | Steve Wolfe | Retrieve previously prepared PCS setup from /home/wolfe/pcs_trees shot#1070719666. Change title of seg1 to indicate +B startup Note that the startup I copied seems to have been from a cryo cold shot. It has pg4 at 32msec. Change it to 20msec. Stare at the seg2 waveforms some more. Zcur is at +.025m, RXL starts at -.015 then sweeps out to -.006 or so around 1sec; not sure what that's about. ZXL is at -.08, out of the vessel, coming back up a little starting at 1.2sec. I understand the low part, not the swing back up to -.06; maybe something about keeping EF2 under control? The active xpoint will be upper. RXU is at -.06, way inside which is the JFT2-M shape. ZXU comes down to -.02 and is flat from .7sec to EOF. I think that makes sense Ip is +900kA, that's correct. EF2_by_EF4 is at +4000 and IC_EF4 is -2000. That looks very strange. Better check it. I guess it's right. The original (1060503022) has EF2_by_EF4 and IC_EF4 in opposite directions, and the actual shot has EF4 near zero, EF2_U near -5000 and EF2_L near -1000. These should reverse. I expect I'll have trouble with these guys railing today. Why would anybody want to run this weird shape? Ah, found an error. Jim asked for 5.4T, but the original must have been higher because the TF program is at 155600. Change to 150000. There is fairly big puff on PG3 in seg2, around 65V; it starts the seg between 30 and 40V. That looks wrong. On the real shot it shuts off during the H-mode, but it still doesn't seem to be a good thing, especially since we're asking for a low density. I'm going to drop it. For reference, the original values are (.1,40),(.211,62.59),(.936,67.48), (1.4,10.969),(1.431,-28.4980),(1.94,-30.304), then back to roughly zero after 2sec. Copy a more normal looking puff, early spike then down to zero, from seg3. Confirm that the nl04 feedback gain is normal (9->8) Fix slight negative value on Acoil_cur in seg1 (-.03A, before supply is on). This would do exactly nothing, but it's unasthetic. Load it. Load at 19-Jul-2007 08:06:29.00 Open tree /home/wolfe/pcs_scratch -1 Open tree done | |
| Jul 19 2007 09:19:00:283AM | 1070719001 | Steve Wolfe | Shot#1: Contrary to PO_Plan, ask for B-side lower enabled on first shot. TF resistance is around -153 again today, need to watch that the cooling keeps up. Early disruption, .53sec, EF1L crowbarred at .38sec, PF breaker pulled. Lots of baseline offset messages, probably from the crowbar. Big swing on EF1 and other demands at 0.3sec. EF1L demand goes to -275, which may be more than it can handle; could be part of the reason for the crowbar? There is a major gain switch on ZXL at 0.3sec, from 2 to 6. That's the reason for the jump. I can back off on that a bit. Next: back off on RXL gain switch |
| Jul 19 2007 09:37:31:243AM | 1070719001 | Steve Wolfe | Shot#1: Contrary to PO_Plan, ask for B-side lower enabled on first shot. TF resistance is around -153 again today, need to watch that the cooling keeps up. Early disruption, .53sec, EF1L crowbarred at .38sec, Other supplies (OH's) crowbar around .51 before disruption; PF breaker pulled. Lots of baseline offset messages, probably from the crowbar. Big swing on EF1 and other demands at 0.3sec. EF1L demand goes to -275, which may be more than it can handle; could be part of the reason for the crowbar? There is a major gain switch on ZXL at 0.3sec, from 2 to 6. That's the reason for the jump. I can back off on that a bit. Next: back off on ZXL gain switch There will be a delay while fuses get replaced |
| Jul 19 2007 09:59:31:487AM | 1070719002 | Steve Wolfe | Shot#2: back off on ZXL gain switch at 0.3sec from 6 to 4, add a second switch at .5sec up to P=6. Drop the pg4 form 20 to 18ms #1 was sort of doing what it should until the crowbar. Try again with the stepped gain, when the fuses get replaced. plasma to .9sec, power supplies survived. EF2U goes to zero at 0.6sec H-modes on and off. Lose it vertically presumably because of EF2. Inner gap is too small at 1 to 1.5cm, outer is also a little small. Need to increase clearin. Does that help or hurt EF2U? Note that EF2L is near the opposite rail. |
| Jul 19 2007 10:08:43:247AM | 1070719003 | Steve Wolfe | Shot#3: CLEARIN from .001 to .008 pg4 from 18 to 17ms (disruption tweak) fizzle. pressure was up a little, not enough of a tweak I guess. Drop it another msec and try again. |
| Jul 19 2007 10:42:36:847AM | 1070719004 | Steve Wolfe | Shot#4: Pg4 from 17 to 16ms MRL had to restart event server Scopes and such won't update unless restarted. plasma with Hmodes, lasted full length. ECE ion gauge is off, gate valve didn't open. Need a cell access to fix (Frank and Ron) This shot looks more or less like we want. Gaps still a little bit low. EF2U is railed at zero throughout; not good. RF not too happy yet. next: increase clearin again |
| Jul 19 2007 11:11:45:193AM | 1070719005 | Steve Wolfe | Shot#5: Clearin from .008 to .012, rcur out by 1.5 to 2mm disrupts early again. Breaker pulled. EF1L crowbars at .429sec, similar to shot#1 Nothing obviously wrong with command, seems to be in range Maybe still a little high at 250V, but it shouldn't break, it should just saturate. Current is low and heading down, not near a rail. EF2U again goes to zero later, but can't tell what it would have done if EF1 stayed on. ECE gate valve didn't open; had a fault on Eng screen, reset now Next: back off on ZXL gain some more, hope to keep EF1 on |
| Jul 19 2007 11:54:42:993AM | 1070719006 | Steve Wolfe | Shot#6: Reduce ZXL gain to 3 and 5 for 2nd and 3rd steps Jim notes that upper triangularity is too low on #4, but EF2U is already zeroing out. Power room access for the engineers to investigate the EF1L problem; it doesn't look like a program/demand problem. Didn't find anything. We've had a long delay and pg4 is already low, so no post-diruption tweak this time. Still trying for the bigger clearin. Disrupted later, around .98sec again. EF2U still railing at zero. Gaps were better while they lasted. And we've got a coax resistance warning. OH1 and OH2U. It's double swinging ok, don't know what else to do except not run this shape. The problem is probably the crowbars; it won't double swing on those. I can push the EF4 current more positive, but EF2L will probably rail. |
| Jul 19 2007 11:57:50:543AM | 1070719007 | Steve Wolfe | Shot#7: IC_EF4 from -2000 to -1000 Move early RXL from -.015 to -.006 fizzle. Not surprised, but I'll drop the pg4 one more time. |
| Jul 19 2007 12:16:04:450PM | 1070719008 | Steve Wolfe | Shot#8: Pg4 from 16 to 15ms (one shot too late) IC_EF4 from -2000 to -1000 Move early RXL from -.015 to -.006 Early disruption ef1L crowbars again self-power circ fired, aux power fail, comm fail. Bill needs to look at this again. |
| Jul 19 2007 12:58:22:133PM | 1070719009 | Steve Wolfe | Shot#9: Set up a test EF1L pulse with a double swing under PCS control. Use seg 1, just change gain and waveform on EF1L; turn off seg2. Will need to set fizzle detector on to 1.6sec or so. I'll need to remember to import seg1 from shot 8 to go back to plasma. And turn seg 2 back on. No crowbar but still some oscillations on forward and reverse bridge before and after crossover, as on all plasma shots today. The EF1L bus voltage is extremely noisy, can't really see what's happening. We'll take at least one more test like this. |
| Jul 19 2007 01:24:25:257PM | 1070719010 | Steve Wolfe | Shot#10: EF1L test under dpcs control, with iso-amp boards on ef1 bus voltage swapped so we can see the ef1l bus voltage (we hope) Burst of noise vsible on bus volts this time, near zero current. Next: swapped circ. current regulator boards |
| Jul 19 2007 01:56:55:763PM | 1070719011 | Steve Wolfe | Shot#11: EF1L test under dpcs control, with swapped circ. current regulator boards, since EF1U doesn't seem to have the problem and that regulator is the suspect (until proven innocent). No pulse, breaker opened. comm fail indication There were 24 failed fuses. Something bad happened on that test! Probably unrelated to the original problem, too. |
| Jul 19 2007 03:41:11:290PM | 1070719012 | Steve Wolfe | Shot#12: EF1L test under PLC control with breakers open. Trying to recover from previous testing disaster. Slew boards back where they belong. Reverse bridge gates removed. Control signals looked correct on this test. No -9V signals from slew boards in either supply. |
| Jul 19 2007 03:57:22:997PM | 1070719013 | Steve Wolfe | Shot#13: EF1L test under PLC control with breakers open. Trying to recover from previous testing disaster. Reverse bridges in. Before this test, Bill and Gary confirmed that swapping the slew boards does not work, it causes the bad control voltages that resulted in the fusicide on shot 11. No power test looks ok. Next: try with breakers closed. |
| Jul 19 2007 04:05:56:460PM | 1070719014 | Steve Wolfe | Shot#14: EF1L forward bridge test under PLC control. test ok. |
| Jul 19 2007 04:22:49:260PM | 1070719015 | Steve Wolfe | Shot#15: EF1L reverse bridge test under PLC control. test ok. Next ask for + bridge which isn't connected, should get nothing. |
| Jul 19 2007 04:30:22:970PM | 1070719016 | Steve Wolfe | Shot#16: EF1L test: ask for + bridge which isn't connected, should get nothing. And did get nothing. Next plug both bridges in and try a forward bridge test. |
| Jul 19 2007 04:45:13:113PM | 1070719017 | Steve Wolfe | Shot#17: plug both bridges in and try a forward bridge test. test ok. Next: back to the double swing shot, like #10. |
| Jul 19 2007 04:55:21:500PM | 1070719018 | Steve Wolfe | Shot#18: back to the double swing shot, like #10. Double swing ok, Loks like the pathology on the voltage and circulating current is still there. Next: test EF1U, just because it's had boards in and out. |
| Jul 19 2007 05:15:46:837PM | 1070719019 | Steve Wolfe | Shot#19: test EF1U, just because it's had boards in and out. Copy double swing waveform from EF1L, and load gains from wire 12 (EF1L). test ran ok. On this test the EF1U has a similar glitchiness near crossover as the EF1L. This implies that the problem that was causing the self-power circuit firing and crowbars on EF1L is a "feature". Looked back at 1060503022 and it has a similar burst on the EF1U voltage, just not as big as today's. |
| Engineering Operator Comments | ||||
| Shot | Time | Type | Status | Comment |
| 1 | 09:05:30:940AM | Plasma | Bad | EF1 self powered 1Fuse lower |
| 2 | 09:51:22:540AM | Plasma | Ok | |
| 3 | 10:04:03:227AM | Plasma | Ok | |
| 4 | 10:31:24:133AM | Plasma | Ok | |
| 5 | 11:00:42:197AM | Plasma | Bad | EF1 Fault |
| 6 | 11:39:27:477AM | Plasma | Ok | |
| 7 | 11:55:32:243AM | Plasma | Ok | |
| 8 | 12:08:22:133PM | Plasma | Bad | Type fault self powered circuit fired iired |
| 9 | 12:45:15:837PM | Test | Ok | EF1 powersupply test shot Hybrid Control |
| 10 | 01:07:02:113PM | Test | Ok | EF1L power supply test shot Hybrid Control |
| 11 | 01:25:58:167PM | Test | Bad | EF1 power supply type 1 fault |
| 12 | 03:34:23:953PM | Test | Ok | EF1 no power test |
| 13 | 03:53:10:990PM | Test | Ok | EF1L No Power Test |
| 14 | 04:02:10:923PM | Test | Ok | EF1 power test |
| 15 | 04:17:35:363PM | Test | Ok | EF1L Reverse PLC Test |
| 16 | 04:26:39:843PM | Test | Ok | EF1 Test |
| 17 | 04:41:24:437PM | Test | Ok | EF1 power test |
| 18 | 04:50:08:727PM | Test | Ok | EF1L double swing |
| 19 | 04:59:20:053PM | Test | Ok | EF1U Test |
| System Availability | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Jul 19 2007 09:05:06:643AM | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Jul 19 2007 04:49:48:290PM | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||