| Miniproposals | ||||||||||
|
| Operators | |
| Session leader(s): | Joe Daigle |
| Physics operator(s): | Steve Wolfe |
| Engineering operator(s): | Vinny Bertolino,Joe Daigle |
| Engineering Operator Run Comment |
| Rice/Wolfe |
| Session Leader Plans |
| Physics Operators Plans |
| Session Leader Summaries |
| Physics Operator Summaries |
| Entered: Jul 7 2004 04:36:55:160PM |
| Author: To Be Determined |
| Phys op summary for 951102 SL: Daigle PO: Wolfe EO: Daigle/Bertolino This "run" was in support of power supply debugging for the EF4 single-phasing problem. Some progress was made, despite the fact that we managed to scare ourselves (again) over some high readings on the OH coax instrumentation. The interesting thing about this run, from the operator stanpoint, is that we developed a power supply testing segment in which the currents are set up to match the evolution on a plasma shot with the feedback on the coil currents rather than plasma parameters. This can be used to investigate power system effects that depend on having the coils act as they do with plasma, without needing to actually make plasmas. Shot 951102015 is a good starting place for using this. Segment 1 was set up to feedback on coil currents. The other segments are turned off. P waveforms were cut in from the traces for shot 951031020. In the course of the original setup of this segment, I tickled at least two PCS bugs, described below. I also made several operator errors which had to be found and identified over the first four shots. Details are in the logbook, under the PHYSICS_OPERATOR topic. Once the programming was set up as desired, shot 5 exhibited the problematic EF4 behavior. The cause is apparently related to the fact that during the early part of the flux swing (0-300msec) there is a pause in the drop of the EF4 current because of induced voltage from the other coils. This pause is uncomfortably close to the 750A level at which the EF4 circuit begins integrating the voltage to determine when to block gates. The effect shows up less robustly when tests are run without the other coils. Starting on shot 5, we went through a mild panic, with a bit of deja vu, when we started getting alarm messages from the OH coax resistance monitor. We ran a series of tests, at 10kA and 5kA and back to 10kA, interspersed with DC measurements, over the course of which the OH2U readings went up to around 2uohm and then seemed to stabilize at that level. The OH2L readings also rose up to 0.98uohm with the same trend characteristics. We believe that this effect, which is similar to what was observed last campaign on the OH2L reading, is either due to faulty instrumentation or a strange cooling system effect. It was noted that the TC readings associated with the coax resistance measurements actually go down from the start of the shot to the end. More reasonable behavior (rising temperature) was observed on the OH1 coax, and had been the case on the OH2U last June. In any case, since the situation did not seem to be deteriorating any further, we resumed testing at full currents on shot 13. The readings then came down to 1.57uohm by shot 15. The run was stopped after #15 to allow the engineers to work on a fix to the EF4 supply, based on the information obtained so far. The plan for fixing the EF4 problem is to modify the circuitry to filter the shunt signal better and try to take action starting at a lower current level. The new circuitry will be ready for tomorrow, when we will try again. As mentioned, two PCS bugs were uncovered in the course of setting up the new segment programming. Both may or may not be related to the fact that I started by zero'ing the entire segment. 1. On wires that started out fresh from having been zero'ed, I had difficulty getting the main display to show the name in upper case, indicating an active channel, even after putting in finite P gain and setting a non-zero controller. I finally succeeded in getting it to show up correctly by (?) repeatedly bringing up the edit_wire screen and the controller screen and making minor tweaks to try to force an action, intermingled with some build segments. I don't know what finally made it work, but I eventually managed to get them all UPCASED. I also don't know if this would have mattered to anything but the screen display. 2. Trying to bring up the controller screen for wires with initially null controllers caused PCS to bomb back to the IDL prompt. I somehow got around this on wires 8 and 10 (IC_OH2L and IC_OH2U), possibly by using copy wire to get something into the controllers and then editing that. I couldn't manage to get it to work at all on wire 16 (which I was trying to use for IC_EF3) but I did manage to put this on wire 7 successfully; I'm still not sure how. The sequence that finds the bug was: a) Start with a zeroed wire and bring up the edit_wire screen b) Put in a valid name and supply predictor values and APPLY, setting the FACTOR when the window appears. c) Push the CONTROLLER button d) Acknowledge the message that it will start with a default controller e) Watch it blow up. The first error is in TREE_M at line 462, called by EDIT_CONTROLLER at line 693. |
| Session Leader Comments |
| Physics Operator Comments | |||
| Nov 2 1995 09:06:59:900AM | 951102001 | Steve Wolfe | Shot 951102002
In aid of poweer supply tests, set up a segment 1 with current waveforms from 951031020. Had considerable trouble setting this up in PCS. Aside from mechanics, the following bugs were noted: It took several iterations of calling up edit_wire and the controller screen to get a newly named wire to turn upper case on the display. On several wires with no initial controller, PCS bombed out after trying to go to the controller screen (push controller, acknowledge widget about none present, bomb) I never got wire 16 to work. No supplies : Hybrid was turned off! Also electrons, and some others. |
| Nov 2 1995 09:13:38:920AM | 951102002 | Steve Wolfe | Shot 951102002
In aid of poweer supply tests, set up a segment 1 with current waveforms from 951031020. Had considerable trouble setting this up in PCS. Aside from mechanics, the following bugs were noted: It took several iterations of calling up edit_wire and the controller screen to get a newly named wire to turn upper case on the display. On several wires with no initial controller, PCS bombed out after trying to go to the controller screen (push controller, acknowledge widget about none present, bomb) I never got wire 16 to work. Trees: HYbrid,MHD, Electrons and MHD on, Spectroscopy left off ONly get hybrid command after 0.1 sec. Operator error - had the on time from segment 2 (0.1 sec) |
| Nov 2 1995 09:45:25:820AM | 951102003 | Steve Wolfe | Shot 951102003
And try again: In aid of poweer supply tests, set up a segment 1 with current waveforms from 951031020. Had considerable trouble setting this up in PCS. Aside from mechanics, the following bugs were noted: It took several iterations of calling up edit_wire and the controller screen to get a newly named wire to turn upper case on the display. On several wires with no initial controller, PCS bombed out after trying to go to the controller screen (push controller, acknowledge widget about none present, bomb) I never got wire 16 to work. Trees: HYbrid,MHD, Electrons and MHD on, Spectroscopy left off OH's went to their rails. More operator error: I had the predictors turning on at 0 instead of -2. |
| Nov 2 1995 09:53:55:040AM | 951102004 | Steve Wolfe | Shot 951102004
And again: In aid of power supply tests, set up a segment 1 with current waveforms from 951031020. That looked ok, except that EF4 is overcharging because I've got a big voltage waveform in. |
| Nov 2 1995 10:44:30:580AM | 951102005 | Steve Wolfe | Shot 951102005
Back initial voltage to -500V on EF4. Extend FD to .5 sec Some double swing on OH's. EF4 single phased,still not down to nominal OH2U coax resistance is up to 1.3uohm. We may be in trouble. |
| Nov 2 1995 11:38:24:620AM | 951102006 | Steve Wolfe | Shot 951102006
Same programming, they extend integration time on blocking circuit. Still single phased. Coax is up to 1.6uohm. Measured coax directly (through 100* amplifier at 100A current) and get 2uohm. |
| Nov 2 1995 12:07:56:230PM | 951102007 | Steve Wolfe | Shot 951102007
Set up for "Tuesday test" 10kA OH PLC test. Got 2.2uohm. Whatever it is, its getting worse. The TC temperature appears to drop from the beginning of the oulse to the end. This would be consistent with the TC being loose in the coax (wishful thinking?). We will consult with Ian before proceeding. |
| Nov 2 1995 01:27:07:350PM | 951102008 | Steve Wolfe | Shot 951102008
Set up for "Tuesday test" but at 5kA OH PLC test. Resistance measurements on the TC to TC show about 250 ohms for all coaxes. It isn't open. 1.9uohm on OH2U, 0.98 on OH2l They seem to be trending up together. |
| Nov 2 1995 01:34:12:280PM | 951102009 | Steve Wolfe | Shot 951102009
Repeat at 5kA OH PLC test. Will keep doing these to see if there is any trend for a few shots. 2.1 on upper. 0.96 on lower. |
| Nov 2 1995 01:38:30:280PM | 951102010 | Steve Wolfe | Shot 951102010
Repeat at 5kA OH PLC test. Will keep doing these to see if there is any trend for a few shots. upper 2.0 lower .95 |
| Nov 2 1995 01:43:40:680PM | 951102011 | Steve Wolfe | Shot 951102011
Back to 10kA OH PLC test. upper 2.0 lower .95 2.1 and .92 |
| Nov 2 1995 01:44:26:810PM | 951102011 | Steve Wolfe | Shot 951102011
Back to 10kA OH PLC test. upper 2.1 lower .92 |
| Nov 2 1995 01:49:49:710PM | 951102012 | Steve Wolfe | Shot 951102012
Repeat 10kA OH PLC test. upper 2.08 lower .92 |
| Nov 2 1995 02:14:46:830PM | 951102013 | Steve Wolfe | Shot 951102013
Whatever is going on with the coaxes seems to be stabilized. We will go back to full power tests, monitoring the value every shot. No EF4? CAMAC pulse missing? OH2U 1.7 |
| Nov 2 1995 02:17:43:400PM | 951102014 | Steve Wolfe | Shot 951102014
Repeat EF4 went ok that time. oh1: 0.1783 uohm oh2u: 1.6166 uohm oh2l: 0.7622 uohm |
| Nov 2 1995 02:26:37:890PM | 951102015 | Steve Wolfe | Shot 951102015
Set FD back to 0.5sec Looks like EF4 S-p'ed again. oh1: 0.1868 uohm oh2u: 1.5689 uohm oh2l: 0.7640 uohm It may be coming back down??? |
| Engineering Operator Comments | ||||
| Shot | Time | Type | Status | Comment |
| 1 | 08:58:32:980AM | Test | Ok | |
| 2 | 09:07:46:420AM | Test | Ok | |
| 3 | 09:16:22:160AM | Test | Ok | |
| 4 | 09:48:07:550AM | Test | Ok | |
| 5 | 10:02:37:890AM | Test | Ok | |
| 6 | 10:42:38:990AM | Test | Ok | |
| 7 | 11:44:25:110AM | Test | Ok | |
| 8 | 01:20:58:940PM | Test | Ok | |
| 9 | 01:30:07:410PM | Test | Ok | |
| 10 | 01:35:07:770PM | Test | Ok | |
| 11 | 01:40:47:000PM | Test | Ok | |
| 12 | 01:46:48:330PM | Test | Ok | |
| 13 | 02:03:04:390PM | Test | Ok | |
| 14 | 02:13:33:570PM | Test | Ok | |
| 15 | 02:22:51:100PM | Test | Ok | |