24 Dec 2017 UT: E/D 1st half

OSA: G Bechetti

Observers: O Kuhn, A Conrad (1st half) and A Rakich (by phone, 1st half)

Instrument Support: E Solheid

Cloudy at sunset. Delaying opening because clouds are thick.

Configured for LUCI2 (DX) + IRTC (SX)

Summary: Cloudy conditions delayed opening until 8pm, but even then ~5 magnitudes of extinction although these eventually thinned and it was pretty clear just before handover.
  • closed dome test of NS tracking on a fast moving target with changing rates to look for jumps. Targets tried: A/2017 U1 (no longer very fast...), 2017 OT18 (70\x94/hr, -40\x94/hr) and 2017 OT18 (166\x94/hr 273\x94/hr) Logs may show something but a better test could be on a fake ephemeris with yet fater rates. Sometimes the new NS target did not get picked up until PCS was restarted (see IT 7007)
  • open dome test to send a NS AO preset with target=AOref star. AO system did not get the AOref star magnitude (see IT 7011).
  • NSIGUI does not find some targets (see IT 7008; workaround is to say ASTNAME=Iris)
  • wobble data collection not possible - the fastlink connection from DX FLAO to DX AdSec was down, restored the next day.
  • LBC TMS testing done, basically just tried the software on-sky but there not time for much.

~00:00 My NX from rm507-5 to obs4 became unusable. This was after I tried a window capture using gimp. But, not to lay blame on gimp, Geno told me that at that same time his command to authorize IRTC was very, very slow. Once Geno stopped GCS (had to kill it on one side), authorization was quick, but my NX was still unusuable. Suspended and reopened it.

00:32: Monocular LUCI2. Send test hijack preset to A/2017 U1. We are tracking now, but unsure whether it is tracking smoothly or jumping. I hope one can tell from telemetry or the logs, but Al send email to Petr and John to ask. (00:55 --- I paused script after the first item, so as just to send the preset, configure and taken the first image we have still been NS tracking. I put in a blind filter just now).

00:54: More LBC biases. Now LBCB biases look ok. They were noisy earlier this evening (Tamb~ 2 but of course was higher earlier than it is now.)

01:10: We continue to track. Petr confirmed that we need not turn on anything - the info is in the logs. But it would be good to try an object with quickly changing rates near zenith and another at low elevation.

But, A/2017 U1 has relative slow apparent rates now\x85 it\x92s not a good test. We\x92re trying a faster target, 2017 YE is good but 2nd half object (1000\x94 / hr and very variable). We\x92ll try 2017 OT18 and then 2017 NU6.

01:33: Closed dome track preset to NS target 2017 OT18 (70\x94/hr, -40\x94/hr). This time we will have taken about 30-sec at the initial position, and then made 2 dithers. After that, I paused the script (~01:35UT) to allow smooth (we hope) NS tracking with no more dithers. 2009 XP2

01:44: Slew to 2009 XP2. Faster than 2017 OT18. Rates are 166\x94/hr 273\x94/hr. Zenith distance 29 deg. HA = -22:48:41 (hmmm\x85) Double-hmmm\x85. The coordinates reported by PCS do not agree with the ephemeris generated by NSIGUI nor do they agree with the JPL Horizons for this target. We can check LUCI2 image headers for the other targets. Indeed, all headers show same telra, teldec \x85 23:26:33 +08:34:59

02:02: Geno sending a sidereal preset. But he says LUCI2 is not authorized and cannot send a preset even from IRAF (error with IRAF found\x85 telside was set to \x93both\x94 but should be \x93right\x94).

02:06: Sidereal preset successful.

02:08: Set hijack for 2009 XP2 and Geno sending preset. Still, RA/DEC are 23:25 +08:35. I initially had NSIGUI open through NX to obs4. Then when that became unusable, I ssh\x92d to obs4 and opened NSIGUI again. Maybe I should not have done that\x85 open 2 on the same machine. I did kill the first one, after I resumed NX. Will close both NSIGUIs and restart one on the NX to obs4.

02:19: Reopened only one NSIGUI, clicked set to set to 2009 XP2, Geno sent preset, but target coord did not change. Slew was nearly instantaneous and still PCS shows we are at 23:26 08:35 (not at 2009 XP2 which should be at 00:25 08:29). These coordinates match the coordinates of the first NS target, A/2017 U1.

02:24 Geno cancels preset, stops PCS. Sends sidereal preset to star @ 23:28, 09:25.

I clicked set. Geno canceled sidereal preset and sending it again. Now the hijack is correct --- telescope went to 00:25 08:29. Restarting PCS did the trick!

02:39: Geno canceled preset. We\x92re going to open. Next up, 7 Iris.

03:00: Open and collimated. Extinction about 5 magnitudes from clouds.

03:20: Slew to 20_Massalia with a hacked script. Copied 7_IRIS_UTDec24_L2.xml to 20_Massalia_UTDec24_L2.xml and edited the target name in 2 places and the ephemeris filename in one (it occurred only once). Slew is being done. PCS got the non-sidereal. Geno collimated earlier on a nearby star so I did not send a \x93collimate\x94 script.

Note: While trying to edit the program 7IRIS_OT.xml, I got a slew of java errors: java.lang.NullPointerException. I tried loading the program into both the OT on the rm507-5 (v 2017B.1.1.0) and on my laptop (v 2017B.1.2.0). (This was because the program had been created using the ARGOS version of the OT but was opened using 2017B1.1.0 (not ARGOS))

03:23:23 --- LUCI messages show error. Preset failed. We\x92re calling Doug\x85 maybe it is AO related (yes, turns out that the wavefront sensor was not on). As far as the NS preset - FACSUM shows we are on-target and tracking non-sidereally.

03:42: Sending new preset to 20_Massalia. Preset failed - star too bright (did we edit the magnitude in the script?). RA, Dec 00:25 08:xx. Did the hijack try to go to 2009 XP2? We\x92ll repeat after Geno restarted PCS. Now when we run the script, we are going to the correct place.

IIF data: right target & right AO star \x85 thinks mag=0.0. Looks like it is not getting the magnitude from the ephemeris. Script I ran was rm507-5:/home/observer/Service_Observing/7IRIS/20_Massalia_UT24Dec_L2.xml

04:01: Geno had to check pointing. The 20 Massalia is ~8 magnitudes but we didn\x92t see it either in the LUCI AGw1 acq or the AO camera. Pointing corrected.

04:03: Try again. Going to correct place. With pointing corrected, there\x92s a star on CCD47. But still it complains that it\x92s too bright \x930.0 mag\x94. 8-9\x94 off-axis but there.

--- Going to try to collect wobble data...

04:15: WFS arbitrator \x93Not live\x94. Not getting frames to the AdSec. Could be a fiber problem? No - reconstructor is not bin-matched to actual binning used. Bin 1 but reconstructor for bin 2. That was at least one problem, but Doug is still not able to get slopes to the AdSec.

05:02: Doug will turn off the adSec and restart it to see if that clears the problem sending slopes to the DX AdSec.

05:09: W unit powering back up... Still no slopes. Fiber switch in treehouse ok, powered off/on adsec & W unit. But there was LUCI-AO comm, and that was OK.

Going to try some LBC TMS testing...

05:40: Reconfigured to LBCR. Andrew is on the phone and we are about to start.

06:05: Slew to a collimation field @ 03:50:59.55, 58:57:43.5 (close to 2017 WE2).

06:~~: We realize you must run measurement.py from lbcobs account (does not work from LBTO account).

Also, run dohybrid, /redonly or dofpia, /redonly

Initially, donut did not do its job\x85 was going to need manual corrections, but we had to point out of the wind.

06:25: New field at 8hr 20deg (SE) at elevation 52 deg.

Dohybrid, /redonly. This time donut worked. LBCR is converging.

Andrew wanted to try to converging to a more stringent limit:

Dofpia, /redonly, redlimit=150\x85 not quite converged but ok.

set_tms_offset.py

Tms_loop.pl.. cntrl-C but Andrew reports a jump of 130 mic in Y at this time, also a big RX\x85 why?


06:39: Run dofpia, /redonly (first z4 being sent by fpia is -4902 nm\x85 consistent with big change which Andrew changed). Converged.

06:47: We need to take out -0.8 mm before set_tms_offset. Global offset is -0.3 from collimate model.

Need to run dofpia, c

Change z global offset from -0.3 to 0.5 (06:56)

Run set_tms_offset (06:57)

Run superfoc (skip for time)

Change z global back to -0.3

Run script (120-sec exposures at r)

Mirror hit limits in X and Y. X~1mm and Y~-2.8mm

07:09 Repeating once to confirm behavior.

dohybrid,/redonly

dofpia,/redonly

Note total collimation z = \x85 ah, ran into \x93command already in progress\x94... We don\x92t have time to repeat the previous set of procedures which ended in the mirror limits. But, Andrew has identified some problems with the scripts which he\x92s writing up:

For one, the scripts need to allow pointing changes I think...

Another, the assumption was made that FPIA exits by moving the primary back to focus, but it does not. The PSF right GUI reports the instrument offset is -0.8mm after FPIA finishes. The science scripts take care of moving the instrument offset to the correct focus offset for the filter that they use.

(see log from 20171228 where TMS system and scripts worked)...

07:19 Reconfiguring for MODS binocular. It had cleared, but just now, again, a cloud is moving over us.

Topic revision: r1 - 30 Dec 2017, OlgaKuhn
This site is powered by FoswikiCopyright © by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding Foswiki? Send feedback